# Video Analytics

AI-based software technology that analyzes video streams to identify objects, events, behaviors, and patterns, enabling search, alerts, and automated detection.

This is one page of public article previews, not the complete archive. Follow Next page to continue. Summaries are not the original full articles.

## Online Mistake Detection in Instruction-Conditioned Procedural Video

DevFeed: [Online Mistake Detection in Instruction-Conditioned Procedural Video](<https://devfeed.tech/articles/online-mistake-detection-in-instruction-conditioned-procedural-video-145538.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.25567>)

Author: Amit Feldman, Ilan Shimshoni, Rotem Dror

Published: 2026-11-20T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>)

Topics: [Video Analytics](<https://devfeed.tech/topics/video-analytics.md>), [multimodal-ai](<https://devfeed.tech/topics/multimodal-ai.md>), [image recognition](<https://devfeed.tech/topics/image-recognition.md>)

Tags: [actions](<https://devfeed.tech/tags/actions.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-cv](<https://devfeed.tech/tags/cs-cv.md>), [detection](<https://devfeed.tech/tags/detection.md>), [error](<https://devfeed.tech/tags/error.md>), [error-propagation](<https://devfeed.tech/tags/error-propagation.md>), [evaluation](<https://devfeed.tech/tags/evaluation.md>), [execution](<https://devfeed.tech/tags/execution.md>), [f1](<https://devfeed.tech/tags/f1.md>), [instructions](<https://devfeed.tech/tags/instructions.md>), [localization](<https://devfeed.tech/tags/localization.md>), [mistakes](<https://devfeed.tech/tags/mistakes.md>), [model](<https://devfeed.tech/tags/model.md>), [modular](<https://devfeed.tech/tags/modular.md>), [monitoring](<https://devfeed.tech/tags/monitoring.md>), [offline](<https://devfeed.tech/tags/offline.md>), [online](<https://devfeed.tech/tags/online.md>), [pipeline](<https://devfeed.tech/tags/pipeline.md>), [predictions](<https://devfeed.tech/tags/predictions.md>), [procedural](<https://devfeed.tech/tags/procedural.md>), [protocol](<https://devfeed.tech/tags/protocol.md>), [video](<https://devfeed.tech/tags/video.md>)

### AI overview

The paper studies online mistake detection in egocentric procedural video, conditioning predictions on task instructions, observed video and prior predictions. The system identifies and temporally localizes completed action attempts, then classifies them as correct, modified, slips, corrections or additions; omissions are assessed after the video ends. It compares a modular pipeline for step identification, localization and execution verification with a joint model predicting all outputs at once. After repartitioning EgoPER to include erroneous videos in every split, both approaches achieve similar step-segmentation F1@0.5 (93.6% modular; 92.1% joint). The joint model scores higher on binary mistake detection (84.1% vs. 80.7%), especially for slips (56.4% vs. 37.5%). Component analysis suggests stage-to-stage error propagation contributes to the modular model's lower detection score. On test videos containing mistakes, both approaches outperform released AEM and ESTANet checkpoints under the paper's evaluation protocol.

### Source excerpt

arXiv:2611.25567v1 Announce Type: cross Abstract: An assistant monitoring a person performing a procedure is most useful when it can detect mistakes while there is still time to recover. This requires judging actions against task instructions as the video unfolds, identifying the attempted step, locating it in time, and assessing its execution. Existing online detectors do not use the instruction text, while instruction-conditioned methods operate offline. In this paper, we study online, instruction-conditioned mistake detection in egocentric videos. Using only observed video, task instructions, and prior predictions, our model identifies and localizes completed action attempts and classifies them as correct, modified, slip, corrections, or additions; omitted steps are determined after the video ends. We compare a modular pipeline, with separate step identification, temporal localization, and execution verification, to a joint model that predicts all outputs in a single step. To support supervised learning from mistakes, we repartition EgoPER to include erroneous videos in all splits. The approaches achieve similar step-segmentation F1@0.5 (93.6% and 92.1%), while the joint model achieves higher binary mistake-detection F1@0.5 (84.1% vs. 80.7%), with its largest gain on slips (56.4% vs. 37.5%). Component-level analysis suggests that error propagation between stages contributes to the modular approach's lower detection score. On test videos containing mistakes, both designs outperform released AEM and ESTANet checkpoints under our evaluation protocol.

## CGM: Chaos-Guided Graph Attention and Expert Routing for Autism Stereotypy Detection

DevFeed: [CGM: Chaos-Guided Graph Attention and Expert Routing for Autism Stereotypy Detection](<https://devfeed.tech/articles/cgm-chaos-guided-graph-attention-and-expert-routing-for-autism-stereotypy-detection-145094.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.25730>)

Author: Hafsa Aslam, Yue Li, Saba Aslam, Abdur Rasool, Qiao Pan, Muhammad Rizwan, Ovidiu Bagdasar

Published: 2026-11-20T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Machine Learning](<https://devfeed.tech/sources/arxiv-machine-learning.md>)

Topics: [Video Analytics](<https://devfeed.tech/topics/video-analytics.md>), [Mixture of Experts (MoE)](<https://devfeed.tech/topics/mixture-of-experts-moe.md>)

Tags: [ablation](<https://devfeed.tech/tags/ablation.md>), [amplitude](<https://devfeed.tech/tags/amplitude.md>), [angular](<https://devfeed.tech/tags/angular.md>), [attention](<https://devfeed.tech/tags/attention.md>), [autism](<https://devfeed.tech/tags/autism.md>), [chaos](<https://devfeed.tech/tags/chaos.md>), [cs-cv](<https://devfeed.tech/tags/cs-cv.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [detection](<https://devfeed.tech/tags/detection.md>), [frame](<https://devfeed.tech/tags/frame.md>), [gate](<https://devfeed.tech/tags/gate.md>), [graph](<https://devfeed.tech/tags/graph.md>), [mixture-of-experts](<https://devfeed.tech/tags/mixture-of-experts.md>), [model](<https://devfeed.tech/tags/model.md>), [monitoring](<https://devfeed.tech/tags/monitoring.md>), [pairs](<https://devfeed.tech/tags/pairs.md>), [public](<https://devfeed.tech/tags/public.md>), [recordings](<https://devfeed.tech/tags/recordings.md>), [reliability](<https://devfeed.tech/tags/reliability.md>), [routing](<https://devfeed.tech/tags/routing.md>), [sample](<https://devfeed.tech/tags/sample.md>), [sessions](<https://devfeed.tech/tags/sessions.md>)

### AI overview

CGM uses per-joint, per-frame angular-velocity irregularity to guide skeleton graph attention and route each frame to two of four experts. In the reported cohort, windowed Sample Entropy rose in stereotyped clips while amplitude-based descriptors did not. On 116 clips with subject-grouped five-fold cross-validation, CGM achieved 0.930 balanced accuracy, 0.187 above the graph-attention baseline. On held-out Panoptic Studio sessions, it maintained 0.86 specificity without retraining, versus at most 0.35 for chaos-free baselines. Ablations found no reliable gain from routing without irregularity; chaos modulation alone contributed +0.132. The excerpt also reports a severity index with 0.87 split-half reliability and consistency within children across therapy recordings. These are reported study results, not evidence here of broader clinical validation.

### Source excerpt

arXiv:2611.25730v1 Announce Type: cross Abstract: Skeleton graph attention for autism video analysis weights joints by content alone, giving an erratic wrist the same standing as a still one. Stereotypy, however, is defined by how unpredictably movement unfolds rather than how much of it occurs: in our cohort, every amplitude-based kinematic descriptor falls in stereotyped clips, while windowed Sample Entropy of angular velocity alone rises (Hedges' g = +1.05). We compute this irregularity as a per-joint, per-frame tensor that structurally drives computation. The proposed chaos-guided graph attention and mixture-of-experts model (CGM) couples two mechanisms. First, a learned pairwise term added to the graph-attention logit concentrates attention on irregular joints and re-ranks the remaining edge weight toward jointly irregular pairs; its unconstrained coefficient neither vanishes nor inverts, and zeroing the tensor realigns attention in 94% of clips. Second, a gate conditioned on the same signal routes each frame to two of four experts, whose weights track frame-level irregularity. Under subject-grouped five-fold cross-validation on 116 clips, CGM reaches 0.930 balanced accuracy, exceeding the graph-attention baseline by +0.187; the component ordering replicates on the public MMASD dataset, and on held-out Panoptic Studio sessions, CGM keeps 0.86 specificity without retraining, against at most 0.35 for chaos-free baselines. In ablation, routing without irregularity yields no reliable gain, whereas chaos modulation alone contributes +0.132. A severity index from the same tensor reaches 0.87 split-half reliability, separates the cohorts with Hedges' g = +1.08, and stays consistent within children across therapy recordings, supporting video-only monitoring of stereotypy during intervention.

## Long-Range Video Presentation Attack Detection: A Three-Class Benchmark Across 1-100 m and Two Sensor Types

DevFeed: [Long-Range Video Presentation Attack Detection: A Three-Class Benchmark Across 1-100 m and Two Sensor Types](<https://devfeed.tech/articles/long-range-video-presentation-attack-detection-a-three-class-benchmark-across-1-100-m-and-two-sensor-types-145009.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.25108>)

Author: Lambert Igene, Mohammad Zahir Uddin Chowdhury, Afzal Hossain, Sandip Purnapatra, Peter Johnson, Stephanie Schuckers

Published: 2026-11-20T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Machine Learning](<https://devfeed.tech/sources/arxiv-machine-learning.md>)

Topics: [AI Foundation Models](<https://devfeed.tech/topics/ai-foundation-models.md>), [Benchmark](<https://devfeed.tech/topics/benchmark.md>), [Video Analytics](<https://devfeed.tech/topics/video-analytics.md>)

Tags: [attack](<https://devfeed.tech/tags/attack.md>), [benchmark](<https://devfeed.tech/tags/benchmark.md>), [class](<https://devfeed.tech/tags/class.md>), [cs-cv](<https://devfeed.tech/tags/cs-cv.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [dataset](<https://devfeed.tech/tags/dataset.md>), [detection](<https://devfeed.tech/tags/detection.md>), [efficientnet](<https://devfeed.tech/tags/efficientnet.md>), [presentation](<https://devfeed.tech/tags/presentation.md>), [range](<https://devfeed.tech/tags/range.md>), [sensor](<https://devfeed.tech/tags/sensor.md>), [video](<https://devfeed.tech/tags/video.md>)

### AI overview

The benchmark evaluates video-based presentation attack detection across 1-100 m, using 374,000+ frames from 261 identities captured indoors and outdoors with a DSLR and a mobile sensor. It labels clips as live, spoof, or props (occlusions) and compares MobileNetV3, MobileNet, EfficientNet, and ResNet-50. The DSLR maintains near-perfect spoof classification across distances, while mobile performance drops sharply beyond 50 m; MobileNetV3 classifies props at 55% at 75 m. EfficientNet reports the best overall result: 98.48% accuracy and 2.01% ACER. Tests on unseen attack datasets report 89-94% accuracy. The abstract identifies distance and sensor resolution as important difficulty factors; it says the dataset is planned for public release under an End-User License Agreement.

### Source excerpt

arXiv:2611.25108v1 Announce Type: new Abstract: Face presentation attack detection (PAD) systems deployed in border control and airport surveillance are constrained by a critical assumption: subjects are captured at close range, typically under one meter. No existing benchmark systematically addresses multi-distance capture, diverse environmental conditions, and concurrent spoofing and occlusion-based attacks, leaving deployed models underprepared for real-world scenarios. We introduce a video-based PAD dataset of 374,000+ frames from 261 unique identities collected across five standoff distances (1 m-100 m) under indoor and outdoor conditions using a DSLR camera (C) and a mobile sensor (M), with three-class labels: live, spoof, and props (occlusions); models are trained and evaluated on video clips rather than single frames. We fine-tune and benchmark four CNN architectures: MobileNetV3, MobileNet, EfficientNet, and ResNet-50. The higher-resolution DSLR maintains near-perfect spoof classification rate across all distances, while the lower-resolution mobile sensor degrades sharply beyond 50 m, with props classification rate dropping to 55% at 75 m for MobileNetV3. EfficientNet achieves the best overall performance (98.48% accuracy, ACER 2.01%). Generalization tests on unseen attack datasets yield 89-94% accuracy, validating cross-domain robustness. These results establish capture distance and sensor resolution as critical but underexplored axes of PAD difficulty. The dataset will be publicly released under an End-User License Agreement to support reproducible long-range PAD research.

## Detecting AI-Generated Videos with Spiking Neural Networks

DevFeed: [Detecting AI-Generated Videos with Spiking Neural Networks](<https://devfeed.tech/articles/detecting-ai-generated-videos-with-spiking-neural-networks-145327.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2605.05895>)

Author: Minsuk Jang, Yujin Yang, Hee-Seon Kim, Minseok Son, Younghun Kim, Changick Kim

Published: 2026-11-20T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>)

Topics: [Artificial Intelligence](<https://devfeed.tech/topics/ai.md>), [Video Analytics](<https://devfeed.tech/topics/video-analytics.md>), [Deep learning](<https://devfeed.tech/topics/deep-learning.md>)

Tags: [ai](<https://devfeed.tech/tags/ai.md>), [ai-generated-video](<https://devfeed.tech/tags/ai-generated-video.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-cv](<https://devfeed.tech/tags/cs-cv.md>), [detection](<https://devfeed.tech/tags/detection.md>), [mast](<https://devfeed.tech/tags/mast.md>), [motion](<https://devfeed.tech/tags/motion.md>), [residual](<https://devfeed.tech/tags/residual.md>), [semantic](<https://devfeed.tech/tags/semantic.md>), [sequence](<https://devfeed.tech/tags/sequence.md>), [snn](<https://devfeed.tech/tags/snn.md>), [spiking-neural-networks](<https://devfeed.tech/tags/spiking-neural-networks.md>), [temporal](<https://devfeed.tech/tags/temporal.md>), [video](<https://devfeed.tech/tags/video.md>), [videos](<https://devfeed.tech/tags/videos.md>)

### AI overview

MAST detects AI-generated videos from temporal evidence, addressing the cross-generator performance drop reported for existing approaches. The authors identify smoother pixel-level frame residuals and more compact semantic-feature trajectories in generated clips. They also observe that SNNs fire mainly at object and motion boundaries on fake clips, consistent with sparse temporal artifacts concentrated at moments of change. MAST combines a spike-driven branch that processes multi-channel temporal residuals with a frozen semantic encoder. On GenVideo, it reports 93.14% mean accuracy across 10 unseen generators under strict cross-generator evaluation, matching or surpassing the strongest ANN-based detectors. The supplied abstract does not provide further implementation details or establish performance beyond this benchmark.

### Source excerpt

arXiv:2605.05895v3 Announce Type: replace-cross Abstract: Modern AI-generated videos are photorealistic at the single-frame level, leaving inter-frame dynamics as the main remaining axis for detection. Existing detectors typically handle this temporal evidence in three ways: feeding the full frame sequence to a generic temporal backbone, reducing one dominant temporal cue to fixed video-level descriptors, or comparing temporal features to real-video statistics through a detection metric. These strategies degrade sharply under cross-generator evaluation, where artifact type and timescale vary across generators. On caption-paired benchmark, GenVidBench, we identify two signatures that prior detectors do not jointly exploit: AI-generated videos exhibit smoother frame-to-frame temporal residuals at the pixel level, and more compact trajectories in the semantic feature space, indicating a temporal smoothness gap at both levels. We further observe that, when raw video is fed into a Spiking Neural Networks (SNNs), fake clips elicit firing predominantly at object and motion boundaries, unlike real clips, suggesting that the SNN responds to temporal artifacts localized at edges. These cues are sparse, asynchronous, and concentrated at moments of change, which makes SNNs a natural choice for this task: their event-driven, sparsely-activated dynamics align with the structure of the residual signal in a way that dense ANN backbones do not. Building on this observation, we propose MAST, a detector that processes multi-channel temporal residuals with a spike-driven temporal branch alongside a frozen semantic encoder for cross-generator generalization. On the GenVideo benchmark, MAST achieves 93.14\% mean accuracy across 10 unseen generators under strict cross-generator evaluation, matching or surpassing the strongest ANN-based detectors and demonstrating the practical applicability of SNNs to AI-generated video detection.

## SurgCBVR: A Multi-Domain Benchmark for Content-Based Surgical Video-to-Video Retrieval

DevFeed: [SurgCBVR: A Multi-Domain Benchmark for Content-Based Surgical Video-to-Video Retrieval](<https://devfeed.tech/articles/surgcbvr-a-multi-domain-benchmark-for-content-based-surgical-video-to-video-retrieval-142641.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.23347>)

Author: Mohammad Eslami, Mousa Moradi, Negin Ghamsarian, Nazlee Zebardast, Jennifer L Lindsey, Sophia Y. Wang, Lucia Sobrin, Rishi P Singh, Tobias Elze

Published: 2026-11-19T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>)

Topics: [AI-driven information retrieval](<https://devfeed.tech/topics/ai-driven-information-retrieval.md>), [information retrieval](<https://devfeed.tech/topics/information-retrieval.md>), [Video Analytics](<https://devfeed.tech/topics/video-analytics.md>), [AI Foundation Models](<https://devfeed.tech/topics/ai-foundation-models.md>)

Tags: [abstract](<https://devfeed.tech/tags/abstract.md>), [arxiv](<https://devfeed.tech/tags/arxiv.md>), [based](<https://devfeed.tech/tags/based.md>), [benchmark](<https://devfeed.tech/tags/benchmark.md>), [cases](<https://devfeed.tech/tags/cases.md>), [classes](<https://devfeed.tech/tags/classes.md>), [classifier](<https://devfeed.tech/tags/classifier.md>), [clip](<https://devfeed.tech/tags/clip.md>), [computer](<https://devfeed.tech/tags/computer.md>), [computer-vision](<https://devfeed.tech/tags/computer-vision.md>), [content](<https://devfeed.tech/tags/content.md>), [counts](<https://devfeed.tech/tags/counts.md>), [covering](<https://devfeed.tech/tags/covering.md>), [cross](<https://devfeed.tech/tags/cross.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [dataset](<https://devfeed.tech/tags/dataset.md>), [datasets](<https://devfeed.tech/tags/datasets.md>), [domain](<https://devfeed.tech/tags/domain.md>), [domains](<https://devfeed.tech/tags/domains.md>), [eess-iv](<https://devfeed.tech/tags/eess-iv.md>), [retrieval](<https://devfeed.tech/tags/retrieval.md>), [video](<https://devfeed.tech/tags/video.md>)

### AI overview

SurgCBVR benchmarks surgical video-to-video retrieval across eight public datasets: 679 cases and 76,489 clips from six procedure types in ophthalmic, laparoscopic, robotic-assisted, and colonoscopy domains. Its Scene tier retrieves matching operative states across cases and sites; Workflow retrieves and localizes phase transitions and deviations. Fixed clip lengths and frame counts, plus phase-classifier oracle and duration-only baselines, are used to check for label and timing shortcuts. Evaluation covers six generic and surgical foundation models, two retrieval methods, and contrastive heads, with within-dataset, cross-dataset, and pooled-gallery settings restricted to the same procedure type. Frozen representations barely beat a timing-only baseline, transition retrieval is near chance, and trained heads lose up to half their accuracy across datasets, indicating weak generalization in current representations.

### Source excerpt

arXiv:2611.23347v1 Announce Type: cross Abstract: Hospitals record thousands of hours of surgical video annually, yet these archives remain difficult to search by content. Surgeons cannot readily retrieve similar operations, complications, or workflow events. Existing surgical computer vision primarily produces labels rather than retrieving relevant videos. We introduce SurgCBVR, the first benchmark for content-based video-to-video retrieval in surgery. It integrates eight public datasets across four domains-ophthalmic, laparoscopic, robotic-assisted, and colonoscopy-covering six procedure types, 679 cases, and 76,489 clips (106 hours). SurgCBVR comprises two tiers: Scene, retrieving the same operative state across cases and sites, and Workflow, retrieving and localizing phase transitions and deviations. Scene includes 74,765 clips spanning 40 classes, three temporal granularities, and three playback speeds; Workflow contains 1,724 transition clips. Fixed clip lengths and frame counts, together with phase-classifier oracle and duration-only baselines, help assess label and timing shortcuts. We evaluate six generic and surgical foundation models, two video retrieval methods, and contrastive heads trained on SurgCBVR, using within-dataset, cross-dataset, and pooled-gallery retrieval restricted to the same procedure type. Frozen representations barely outperform a timing-only baseline, transition retrieval approaches chance, and trained heads lose up to half their accuracy across datasets. These findings expose substantial limitations in current video representations and establish SurgCBVR as a benchmark for advancing surgery-aware retrieval.

## APS: Adaptive Partitioning-Based Frame Selection for Long Video Understanding

DevFeed: [APS: Adaptive Partitioning-Based Frame Selection for Long Video Understanding](<https://devfeed.tech/articles/aps-adaptive-partitioning-based-frame-selection-for-long-video-understanding-142753.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.24079>)

Author: Junxiang Wen, Yinxi Zhao, Jiawei Chen, Jingying Hu, Lijun Wang, Zhen Qin, Shuiguang Deng

Published: 2026-11-19T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>)

Topics: [Video Analytics](<https://devfeed.tech/topics/video-analytics.md>)

Tags: [adaptive](<https://devfeed.tech/tags/adaptive.md>), [budget](<https://devfeed.tech/tags/budget.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-cv](<https://devfeed.tech/tags/cs-cv.md>), [dynamic-programming](<https://devfeed.tech/tags/dynamic-programming.md>), [experiments](<https://devfeed.tech/tags/experiments.md>), [fixed](<https://devfeed.tech/tags/fixed.md>), [frame](<https://devfeed.tech/tags/frame.md>), [framework](<https://devfeed.tech/tags/framework.md>), [partitioning](<https://devfeed.tech/tags/partitioning.md>), [query](<https://devfeed.tech/tags/query.md>), [relevance](<https://devfeed.tech/tags/relevance.md>), [sampling](<https://devfeed.tech/tags/sampling.md>), [selection](<https://devfeed.tech/tags/selection.md>), [temporal](<https://devfeed.tech/tags/temporal.md>), [training](<https://devfeed.tech/tags/training.md>), [video](<https://devfeed.tech/tags/video.md>), [vision-language-models](<https://devfeed.tech/tags/vision-language-models.md>)

### AI overview

APS is a training-free method for selecting informative frames from long videos under limited visual budgets. It adapts temporal partitions to query-frame relevance by jointly fitting partition boundaries and local models with complexity regularization and dynamic programming; the resulting partitions guide frame-budget allocation by relevance and temporal extent. On Video-MME, MLVU, and LongVideoBench, the abstract reports consistent improvements across diverse large vision-language model backbones and the highest average gains over uniform sampling among competing methods. Further analyses report generalizability across model scales, similarity encoders, and advanced LVLMs.

### Source excerpt

arXiv:2611.24079v1 Announce Type: cross Abstract: Recent advances in large vision-language models (LVLMs) have significantly improved video understanding capabilities, yet processing long videos remains challenging due to the substantial computational cost of extensive visual inputs. Frame selection offers an efficient alternative, but many existing region-based methods rely on fixed or predefined temporal partitions, which may result in suboptimal temporal regions, compromising subsequent budget allocation and frame selection. In this work, we propose Adaptive Partitioning-Based Frame Selection (APS), a training-free framework that adapts temporal partitions to local variations in query-frame relevance. Specifically, APS formulates temporal partitioning as a complexity-regularized fitting problem, jointly optimizing partition boundaries and local models through dynamic programming. The resulting adaptive temporal partitions guide frame budget allocation based on query relevance and temporal extent, enabling informative frame selection under limited visual budgets. Extensive experiments on Video-MME, MLVU, and LongVideoBench demonstrate that APS consistently improves performance across diverse LVLM backbones, achieving the highest average gains over uniform sampling among competing methods. Further analyses validate the effectiveness of adaptive partitioning and demonstrate its generalizability across model scales, similarity encoders, and advanced LVLMs.

## MonitorVLM:A Vision Language Framework for Safety Violation Detection in Mining Operations

DevFeed: [MonitorVLM:A Vision Language Framework for Safety Violation Detection in Mining Operations](<https://devfeed.tech/articles/monitorvlm-a-vision-language-framework-for-safety-violation-detection-in-mining-operations-142499.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2510.03666>)

Author: Jiang Wu, Sichao Wu, Yinsong Ma, Guangyuan Yu, Haoyuan Xu, Lifang Zheng, Jingliang Duan

Published: 2026-11-19T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>)

Topics: [Video Analytics](<https://devfeed.tech/topics/video-analytics.md>), [image recognition](<https://devfeed.tech/topics/image-recognition.md>)

Tags: [action-recognition](<https://devfeed.tech/tags/action-recognition.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-cv](<https://devfeed.tech/tags/cs-cv.md>), [dataset](<https://devfeed.tech/tags/dataset.md>), [detection](<https://devfeed.tech/tags/detection.md>), [experimental](<https://devfeed.tech/tags/experimental.md>), [filter](<https://devfeed.tech/tags/filter.md>), [framework](<https://devfeed.tech/tags/framework.md>), [industrial](<https://devfeed.tech/tags/industrial.md>), [inference](<https://devfeed.tech/tags/inference.md>), [inspection](<https://devfeed.tech/tags/inspection.md>), [interface](<https://devfeed.tech/tags/interface.md>), [language](<https://devfeed.tech/tags/language.md>), [language-models](<https://devfeed.tech/tags/language-models.md>), [large](<https://devfeed.tech/tags/large.md>), [latency](<https://devfeed.tech/tags/latency.md>), [lightweight](<https://devfeed.tech/tags/lightweight.md>), [mining](<https://devfeed.tech/tags/mining.md>), [models](<https://devfeed.tech/tags/models.md>), [module](<https://devfeed.tech/tags/module.md>), [safety](<https://devfeed.tech/tags/safety.md>), [vision](<https://devfeed.tech/tags/vision.md>)

### AI overview

MonitorVLM detects mining safety violations from surveillance video using a domain-specific dataset of 9,000 VQA samples covering 40 regulations. Its clause filter selects relevant rules and reportedly cuts inference latency by 13.56% while maintaining accuracy; its behavior magnifier enlarges worker regions, adding 3.45% precision and 8.62% recall. Against an unfine-tuned 72B vision-language baseline, the framework reports gains of 22.01% in precision, 34.22% in recall, and 28.37% in F1. A web interface supports automatic violation reports with video timestamps. These are reported abstract results; the excerpt provides no further evaluation details.

### Source excerpt

arXiv:2510.03666v3 Announce Type: replace-cross Abstract: Industrial accidents, particularly in high-risk domains such as surface and underground mining, are frequently caused by unsafe worker behaviors. Traditional manual inspection remains labor-intensive, error-prone, and insufficient for large-scale, dynamic environments, highlighting the urgent need for intelligent and automated safety monitoring. In this paper, we present MonitorVLM, a novel vision--language framework designed to detect safety violations directly from surveillance video streams. MonitorVLM introduces three key innovations: (1) a domain-specific violation dataset comprising 9,000 vision--question--answer (VQA) samples across 40 high-frequency mining regulations, enriched with augmentation and auxiliary detection cues; (2) a clause filter (CF) module that dynamically selects the Top-$K$ most relevant clauses, reducing inference latency by 13.56% while maintaining accuracy; and (3) a behavior magnifier (BM) module that enhances worker regions to improve fine-grained action recognition, yielding additional gains of 3.45% in precision and 8.62% in recall. Experimental results demonstrate that MonitorVLM significantly outperforms baseline vision--language models, achieving improvements of 22.01% in precision, 34.22\% in recall, and 28.37% in F1 score over the 72B unfine-tuned baseline. A lightweight web-based interface further integrates MonitorVLM into practical workflows, enabling automatic violation reporting with video timestamping. This study highlights the potential of multimodal large models to enhance occupational safety monitoring in mining and beyond.

## Action Recognition vs. Detection + Tracking: When to Use Which

DevFeed: [Action Recognition vs. Detection + Tracking: When to Use Which](<https://devfeed.tech/articles/action-recognition-vs-detection-tracking-when-to-use-which-143831.md>)

Original publisher: [Read original article](<https://roboflow.com/blog/when-to-use-action-recognition>)

Author: Erik Kokalj

Published: 2026-11-18T22:22:42Z

Content type: comparison

Language: en

Sources: [Roboflow](<https://devfeed.tech/sources/roboflow-blog.md>)

Topics: [Video Analytics](<https://devfeed.tech/topics/video-analytics.md>), [image recognition](<https://devfeed.tech/topics/image-recognition.md>), [Machine learning](<https://devfeed.tech/topics/machine-learning.md>), [rule-engine](<https://devfeed.tech/topics/rule-engine.md>)

Tags: [action-recognition](<https://devfeed.tech/tags/action-recognition.md>), [basketball](<https://devfeed.tech/tags/basketball.md>), [bytetrack](<https://devfeed.tech/tags/bytetrack.md>), [cosmos](<https://devfeed.tech/tags/cosmos.md>), [cosmos-3](<https://devfeed.tech/tags/cosmos-3.md>), [detection](<https://devfeed.tech/tags/detection.md>), [f1](<https://devfeed.tech/tags/f1.md>), [fine-tuning](<https://devfeed.tech/tags/fine-tuning.md>), [fps](<https://devfeed.tech/tags/fps.md>), [motion](<https://devfeed.tech/tags/motion.md>), [tracking](<https://devfeed.tech/tags/tracking.md>), [video](<https://devfeed.tech/tags/video.md>), [workout](<https://devfeed.tech/tags/workout.md>)

### AI overview

The best approach depends on what the video question asks. Action recognition is suited to brief, subtle actions that require several frames: in the F1 test, Cosmos 3 Edge initially missed a quarter-second jack lift at 4 fps, but at 8 fps it became the strongest pit-stop system. In basketball, it caught all five dribbles, while a detector-based dribble rule caught one; it learned made shots from 13 examples. Detection plus ByteTrack is useful for visible objects and locations: it tracked tire removal better than action recognition, but missed jack lifts because the car barely moved and the crew obscured the jack. For body-pose tasks, a pretrained RF-DETR keypoint model plus simple geometry counted workout reps without training; action recognition often identified the exercise but merged repeated movements into one segment. The excerpt recommends enough labeled clips for action recognition, higher sampling rates for actions lasting a second or less, and adding a counter for reps. Detection and tracking depend on reliable detections, and pose distinctions such as squats versus lunges need keypoints rather than plain boxes.

### Source excerpt

We built action recognition and detection or keypoints + tracking systems for six video tasks, from F1 pit stops to workout reps. Here is when each approach wins.

## Event-Based Dynamic Lip Region Estimation During Speech

DevFeed: [Event-Based Dynamic Lip Region Estimation During Speech](<https://devfeed.tech/articles/event-based-dynamic-lip-region-estimation-during-speech-141843.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.23111>)

Author: Hamze Hammami, Omar Eldardeer, Francesco Rea

Published: 2026-11-18T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Machine Learning](<https://devfeed.tech/sources/arxiv-machine-learning.md>)

Topics: [Speech Processing](<https://devfeed.tech/topics/speech-processing.md>), [Video Analytics](<https://devfeed.tech/topics/video-analytics.md>)

Tags: [cameras](<https://devfeed.tech/tags/cameras.md>), [cnn](<https://devfeed.tech/tags/cnn.md>), [continuous](<https://devfeed.tech/tags/continuous.md>), [cs-cv](<https://devfeed.tech/tags/cs-cv.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [encoder](<https://devfeed.tech/tags/encoder.md>), [event](<https://devfeed.tech/tags/event.md>), [grid](<https://devfeed.tech/tags/grid.md>), [high-dynamic-range](<https://devfeed.tech/tags/high-dynamic-range.md>), [human-robot-interaction](<https://devfeed.tech/tags/human-robot-interaction.md>), [network](<https://devfeed.tech/tags/network.md>), [recurrent](<https://devfeed.tech/tags/recurrent.md>), [region](<https://devfeed.tech/tags/region.md>), [rnn](<https://devfeed.tech/tags/rnn.md>), [roi](<https://devfeed.tech/tags/roi.md>), [sensor](<https://devfeed.tech/tags/sensor.md>), [speaker](<https://devfeed.tech/tags/speaker.md>), [speakers](<https://devfeed.tech/tags/speakers.md>), [speech](<https://devfeed.tech/tags/speech.md>), [stateful](<https://devfeed.tech/tags/stateful.md>), [stateless](<https://devfeed.tech/tags/stateless.md>), [voxel](<https://devfeed.tech/tags/voxel.md>)

### AI overview

The method estimates a dense lip-region heatmap from full event-camera input, addressing the reliance of event-based speech systems on pre-cropped regions or mouth landmarks. It encodes events in polarity-separated voxel grids and uses a two-pathway continuous-time recurrent network: a stateless fast pathway and a stateful slow pathway, both using learned bounded time constants tied to elapsed time. On four unseen speakers it reaches 69.5% on lips and switches targets 2.1 times per second, compared with 8.9 and 11.0 switches for larger convolutional baselines. A spatial encoder improves precision but fails at a new face scale. On a real sensor despite training only on simulated events, the method scores 58.0% versus the CNN's 43.2%, with fewer than half as many target switches.

### Source excerpt

arXiv:2611.23111v1 Announce Type: cross Abstract: Event cameras offer high temporal resolution and low power consumption, making them promising for speech-related applications. These include automatic lip reading, audiovisual speech analysis, and human-robot interaction, particularly in challenging environments characterized by low illumination or high dynamic range. Event-based lip reading, speaker recognition, and voice activity detection, however, mostly operate on a cropped region of interest (ROI), by hand or by the dataset, and no event-based method supplies that ROI as a spatial region; existing event-based face analysis gives the mouth as landmarks, which presents a mismatch between application and localization. The present work predicts a dense heatmap over the lip region from the full event "frame". Events are encoded as a polarity-separated voxel grid and processed at every location by a two-pathway continuous-time recurrent network (CT-RNN) with learned, log-space bounded time constants acting on the real elapsed time of each bin: a stateless fast pathway and a stateful slow pathway, utilised for refining and anchoring the prediction. The model reaches 69.5% on lips on four unseen speakers and switches target only 2.1 times per second, against 8.9 and 11.0 for convolutional baselines with more parameters. Adding a spatial encoder improves precision but fails on a new face scale. On a real event sensor, trained only on simulated events, the proposed model reaches 58.0% on lips against 43.2% for the CNN, with fewer than half its target switches.

## Social Signal Bench: Evidence-Weighted Evaluation of Co-occurring Social Signals in Video and Audio

DevFeed: [Social Signal Bench: Evidence-Weighted Evaluation of Co-occurring Social Signals in Video and Audio](<https://devfeed.tech/articles/social-signal-bench-evidence-weighted-evaluation-of-co-occurring-social-signals-in-video-and-audio-141801.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.22704>)

Author: Siddharth Ravi, Anders R. Bargum, Hugo Anselme, Patrizio Cugia di Sant'Orsola, Alberto Cardaci, Flora Kincso Parragi, Vincent Borowiec, Line H. Clemmensen

Published: 2026-11-18T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Machine Learning](<https://devfeed.tech/sources/arxiv-machine-learning.md>)

Topics: [multimodal-ai](<https://devfeed.tech/topics/multimodal-ai.md>), [AI Models](<https://devfeed.tech/topics/ai-models.md>), [Dataset development](<https://devfeed.tech/topics/dataset-development.md>), [Video Analytics](<https://devfeed.tech/topics/video-analytics.md>)

Tags: [audio](<https://devfeed.tech/tags/audio.md>), [bench](<https://devfeed.tech/tags/bench.md>), [benchmark](<https://devfeed.tech/tags/benchmark.md>), [computer-vision](<https://devfeed.tech/tags/computer-vision.md>), [cs-cv](<https://devfeed.tech/tags/cs-cv.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [engagement](<https://devfeed.tech/tags/engagement.md>), [evaluation](<https://devfeed.tech/tags/evaluation.md>), [frontier-models](<https://devfeed.tech/tags/frontier-models.md>), [recognition](<https://devfeed.tech/tags/recognition.md>), [signal](<https://devfeed.tech/tags/signal.md>), [social](<https://devfeed.tech/tags/social.md>), [video](<https://devfeed.tech/tags/video.md>)

### AI overview

Social Signal Bench evaluates co-occurring conversational signals in 47 hours of video and audio, addressing limits of conventional multi-label F1. Its video dataset covers ten signals in 25 audiovisual hours; its audio dataset covers six signals in 22 hours. Each has a public online-conversation training set, including 10 expert-annotated hours with the remainder pseudo-labeled, and a private expert-labeled test set spanning four domains. The proposed Evidence-Weighted F1 weights reference labels by annotator evidence and gives graded credit to semantically related predictions via optimal per-clip matching; controlled perturbations show it distinguishes related from unrelated errors, unlike conventional F1. Under a shared protocol, post-trained 7B open models outperform general-purpose frontier models on video and match them on audio.

### Source excerpt

arXiv:2611.22704v1 Announce Type: cross Abstract: Recognizing social signals such as interest and hesitation requires interpreting context-dependent behaviors that may co-occur in conversation. Existing affective computing benchmarks rarely evaluate such co-occurring signals, and conventional multi-label F1 overlooks semantic relationships and annotator disagreement. We introduce Social Signal Bench, a 47-hour densely annotated benchmark for the evaluation of social signal recognition and engagement, comprising two datasets: Social Signal Bench Video (25 hours), covering ten social signals in audiovisual recordings, and Social Signal Bench Audio (22 hours), covering six signals in audio-only recordings. Each dataset pairs a public training set of online conversational video, of which 10 hours are expert-annotated and the remainder pseudo-labeled, with a private expert-labeled test set spanning four conversational domains. We also propose Evidence-Weighted F1 (EW-F1), which weights reference labels by annotator evidence and assigns graded credit to semantically related predictions through optimal per-clip matching. Controlled perturbations demonstrate that EW-F1 distinguishes semantically related errors from unrelated ones, unlike conventional F1. Under a shared evaluation protocol, 7B open models post-trained on Social Signal Bench outperform general-purpose frontier models on video and match them on audio.

## DeepfakeBench-MM: A Comprehensive Benchmark for Multimodal Deepfake Detection

DevFeed: [DeepfakeBench-MM: A Comprehensive Benchmark for Multimodal Deepfake Detection](<https://devfeed.tech/articles/deepfakebench-mm-a-comprehensive-benchmark-for-multimodal-deepfake-detection-139224.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2510.22622>)

Author: Kangran Zhao, Xiaoyu Zhang, Yupeng Chen, Yize Chen, Weinan Guan, Bo Zheng, Yufan Liao, Chengzhe Sun, Soumyya Kanti Datta, Qingshan Liu, Siwei Lyu, Baoyuan Wu

Published: 2026-11-17T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Cryptography and Security](<https://devfeed.tech/sources/arxiv-cryptography-and-security.md>)

Topics: [Video Analytics](<https://devfeed.tech/topics/video-analytics.md>), [Dataset development](<https://devfeed.tech/topics/dataset-development.md>), [benchmark overfitting machine learning](<https://devfeed.tech/topics/benchmark-overfitting-machine-learning.md>)

Tags: [benchmark](<https://devfeed.tech/tags/benchmark.md>), [cs-cr](<https://devfeed.tech/tags/cs-cr.md>), [cs-cv](<https://devfeed.tech/tags/cs-cv.md>), [cs-mm](<https://devfeed.tech/tags/cs-mm.md>), [deepfake](<https://devfeed.tech/tags/deepfake.md>), [deepfake-detection](<https://devfeed.tech/tags/deepfake-detection.md>), [multimodal](<https://devfeed.tech/tags/multimodal.md>)

### AI overview

The study introduces Mega-MMDF, a dataset of 100,000 real and 1.2 million forged audio-visual clips spanning 22 manipulation pipelines, and DeepfakeBench-MM, a standardized framework for evaluating 10 detectors across 8 datasets. Training-based detectors achieved their highest mean AUC with Mega-MMDF, gaining up to 0.0784 over each detector's best alternative training dataset. Cross-pipeline results show asymmetric generalization: training on composite forgeries did not always support detection of their component manipulations in isolation. Modality masking increased false alarms and missed detections, while temporal misalignment raised mean predicted forgery probabilities. Code and dataset access are subject to application and usage terms.

### Source excerpt

arXiv:2510.22622v2 Announce Type: replace Abstract: Audio-visual deepfakes can spread misinformation and enable impersonation, undermining trust in digital media. However, existing datasets provide limited coverage of forgery operations and their combinations, constraining detector development and systematic evaluation. We introduce Mega-MMDF, containing 100,000 real and 1.2 million forged clips across 22 pipelines spanning modality-specific manipulations, stacked forgeries, and joint audio-video generation. To assess its training value and enable fair comparisons, we develop DeepfakeBench-MM, which standardizes training and evaluation across 8 datasets and 10 detectors. Training-based detectors achieve their highest mean AUC when trained on Mega-MMDF, with gains of up to 0.0784 over the best alternative training dataset for each detector. Experiments on data scale and pipeline coverage further support the value of large, diverse training data. Cross-pipeline evaluations reveal bidirectional and asymmetric generalization. Training on composite forgeries does not necessarily enable detection of their constituent operations in isolation. Our analysis of altered inputs shows that modality masking can increase false alarms and missed detections, and temporal misalignment can raise mean predicted forgery probabilities. Together, these resources provide a foundation for understanding and improving multimodal deepfake detection, with publicly available code and dataset access subject to application and usage terms.

## WeVisForensics: Self-Evolving Agent for Omni AI-Generated Video Detection

DevFeed: [WeVisForensics: Self-Evolving Agent for Omni AI-Generated Video Detection](<https://devfeed.tech/articles/wevisforensics-self-evolving-agent-for-omni-ai-generated-video-detection-137688.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.21373>)

Author: Yipeng Du, Yuxi Li, Ying Zhang, Chong Sun, Chen Li, Jing Lyu, Ying Tai

Published: 2026-11-17T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>)

Topics: [Multi Agent Systems](<https://devfeed.tech/topics/multi-agent-systems.md>), [agent orchestration](<https://devfeed.tech/topics/agent-orchestration.md>), [Reinforcement learning](<https://devfeed.tech/topics/reinforcement-learning.md>), [supervisor agent pattern](<https://devfeed.tech/topics/supervisor-agent-pattern.md>), [hierarchical agent systems](<https://devfeed.tech/topics/hierarchical-agent-systems.md>), [Video Analytics](<https://devfeed.tech/topics/video-analytics.md>), [multimodal-ai](<https://devfeed.tech/topics/multimodal-ai.md>), [image recognition](<https://devfeed.tech/topics/image-recognition.md>)

Tags: [abstract](<https://devfeed.tech/tags/abstract.md>), [adaptive](<https://devfeed.tech/tags/adaptive.md>), [agent](<https://devfeed.tech/tags/agent.md>), [agent-workflows](<https://devfeed.tech/tags/agent-workflows.md>), [agentic](<https://devfeed.tech/tags/agentic.md>), [agents](<https://devfeed.tech/tags/agents.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-generated-video](<https://devfeed.tech/tags/ai-generated-video.md>), [annotations](<https://devfeed.tech/tags/annotations.md>), [artifacts](<https://devfeed.tech/tags/artifacts.md>), [arxiv](<https://devfeed.tech/tags/arxiv.md>), [audio](<https://devfeed.tech/tags/audio.md>), [benchmarks](<https://devfeed.tech/tags/benchmarks.md>), [call](<https://devfeed.tech/tags/call.md>), [capabilities](<https://devfeed.tech/tags/capabilities.md>), [chain-of-thought](<https://devfeed.tech/tags/chain-of-thought.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-cv](<https://devfeed.tech/tags/cs-cv.md>), [detection](<https://devfeed.tech/tags/detection.md>), [domain](<https://devfeed.tech/tags/domain.md>), [experiments](<https://devfeed.tech/tags/experiments.md>), [video](<https://devfeed.tech/tags/video.md>)

### AI overview

WeVisForensics detects AI-generated videos by coordinating a main investigator with visual and audio forensic inspector agents. The investigator selects which artifact to inspect, the time window, and when to stop; shared feedback also helps refine the inspectors. Its dual-timescale policy optimization trains the investigator faster and improves inspectors more gradually using annotated artifact intervals. ForenCue-30K provides role-specific annotations, interaction trajectories, and a reinforcement-learning pool. The abstract reports 95.2% average detection accuracy across in-domain, out-of-domain, and agentic benchmarks, 9.0% above the strongest prior detector, with verdicts grounded in localized evidence.

### Source excerpt

arXiv:2611.21373v1 Announce Type: cross Abstract: Recent advances in video generation call for reliable detection grounded in explicit forensic evidence. However, current methods primarily depend on single-pass chain-of-thought reasoning that overlooks brief or localized artifacts, or on predefined multi-agent workflows that offer limited flexibility in selecting relevant inspections. Even with adaptive investigation, optimizing the investigation policy alone does not directly improve the forensic capabilities of fixed experts. We introduce WeVisForensics, an agentic framework in which a main investigator agent coordinates a self-evolving library of forensic inspector sub-agents spanning visual and audio cues. The investigator learns which artifact type to examine, over which window, and when to conclude, and these interactions provide shared feedback for both levels of agents. Our dual-timescale policy optimization strategy, DTPO, closes this loop through fast reinforcement learning of the investigator and slower refinement of the inspectors, guided by annotated artifact intervals, to report subtle artifacts they once dismissed. Furthermore, we curate ForenCue-30K, combining role-specific forensic annotations and interaction trajectories to supervise individual inspections and their coordination, with an accompanying reinforcement-learning pool. Experiments show that WeVisForensics averages 95.2% detection accuracy across in-domain, out-of-domain and agentic benchmarks, 9.0% above the strongest prior detector, while grounding its verdicts in localized evidence.

## CreaTrace: Video Creative Plagiarism Detection and Evidence Tracing

DevFeed: [CreaTrace: Video Creative Plagiarism Detection and Evidence Tracing](<https://devfeed.tech/articles/creatrace-video-creative-plagiarism-detection-and-evidence-tracing-137518.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.19976>)

Author: Chang Liu, Sheng Li, Zhenxing Qian, Xinpeng Zhang

Published: 2026-11-17T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>)

Topics: [Video Analytics](<https://devfeed.tech/topics/video-analytics.md>), [image recognition](<https://devfeed.tech/topics/image-recognition.md>), [multimodal-ai](<https://devfeed.tech/topics/multimodal-ai.md>)

Tags: [abstract](<https://devfeed.tech/tags/abstract.md>), [appearance](<https://devfeed.tech/tags/appearance.md>), [camera](<https://devfeed.tech/tags/camera.md>), [consistency](<https://devfeed.tech/tags/consistency.md>), [contrastive-learning](<https://devfeed.tech/tags/contrastive-learning.md>), [copy](<https://devfeed.tech/tags/copy.md>), [copyright](<https://devfeed.tech/tags/copyright.md>), [creative](<https://devfeed.tech/tags/creative.md>), [cross](<https://devfeed.tech/tags/cross.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-cv](<https://devfeed.tech/tags/cs-cv.md>), [dataset](<https://devfeed.tech/tags/dataset.md>), [decoupling](<https://devfeed.tech/tags/decoupling.md>), [design](<https://devfeed.tech/tags/design.md>), [detection](<https://devfeed.tech/tags/detection.md>), [effectiveness](<https://devfeed.tech/tags/effectiveness.md>), [event](<https://devfeed.tech/tags/event.md>), [events](<https://devfeed.tech/tags/events.md>), [experiments](<https://devfeed.tech/tags/experiments.md>), [plagiarism](<https://devfeed.tech/tags/plagiarism.md>), [tracing](<https://devfeed.tech/tags/tracing.md>), [video](<https://devfeed.tech/tags/video.md>)

### AI overview

CreaTrace targets video plagiarism that reproduces creative ideas despite substantial visual differences, a case that visual-similarity-focused copy detection may miss. It first segments long videos into narrative events, then matches candidate event pairs using narrative semantics and camera language. Creative Decoupling separates abstract creative patterns from their visual realizations, while Creative Swap Contrastive Learning uses cross-realization feature swapping to strengthen creative consistency. The framework aims to detect creative similarity, localize matching evidence in time, and produce interpretable reports. The abstract reports effective, robust performance across diverse plagiarism scenarios but provides no detailed metrics here.

### Source excerpt

arXiv:2611.19976v1 Announce Type: cross Abstract: Video creative plagiarism refers to the unauthorized reproduction of creative ideas across videos despite substantial differences in visual appearance, posing a new challenge to video copyright protection. However, existing video copy detection methods primarily focus on visual content similarity, struggling to capture abstract creative that are entangled with specific visual realizations. To address this issue, we introduce a novel task, named Video Creative Plagiarism Detection (VCPD), and propose an iterative creative evidence tracing framework, termed CreaTrace, to verify creative similarity and accurately localize the corresponding evidence segments. Specifically, we design a Multimodal Semantic Event Segmentation strategy to divide similar long videos into coherent narrative events, facilitating event-level dataset construction and retrieving similar segments. Based on these events, CreaTrace employs Coarse Creative Matching to identify candidate event pairs through narrative semantics and camera language. Then CreaTrace presents Creative Decoupling to disentangle abstract creative patterns from specific visual realizations. Meanwhile, Creative Swap Contrastive Learning is devised to enhance creative consistency through cross-realization feature swapping. Benefiting from the above designs, our CreaTrace enables accurate creative similarity detection and temporal localization, while providing interpretable evidence report. Extensive experiments demonstrate the effectiveness and robustness of CreaTrace across diverse creative plagiarism scenarios.

## Adaptive Distance Detection and Adaptive Instance Segmentation

DevFeed: [Adaptive Distance Detection and Adaptive Instance Segmentation](<https://devfeed.tech/articles/adaptive-distance-detection-and-adaptive-instance-segmentation-133060.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.15991>)

Author: Kai Yi Ji

Published: 2026-11-13T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Machine Learning](<https://devfeed.tech/sources/arxiv-machine-learning.md>)

Topics: [Video Analytics](<https://devfeed.tech/topics/video-analytics.md>)

Tags: [adaptive](<https://devfeed.tech/tags/adaptive.md>), [algorithm](<https://devfeed.tech/tags/algorithm.md>), [approach](<https://devfeed.tech/tags/approach.md>), [combines](<https://devfeed.tech/tags/combines.md>), [cs-cv](<https://devfeed.tech/tags/cs-cv.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [detection](<https://devfeed.tech/tags/detection.md>), [model](<https://devfeed.tech/tags/model.md>), [object-detection](<https://devfeed.tech/tags/object-detection.md>), [semantic](<https://devfeed.tech/tags/semantic.md>)

### AI overview

In Counter Strike: Condition Zero enemy detection, the authors compare a customized semantic segmentation model with an object detector trained using YOLOv5s. They report that object detection is slightly faster, while semantic segmentation better captures an object's exact shape; they say the inference-time gap may grow as both models become more complex. They propose Adaptive distance detection to combine the two approaches and Adaptive instance segmentation to combine semantic and panoptic segmentation. The excerpt gives no details or evaluation results for these proposed methods.

### Source excerpt

arXiv:2611.15991v1 Announce Type: cross Abstract: This paper is about object detection model and semantic segmentation model in an FPS game Counter Strike: Condition Zero to detect enemies. Semantic segmentation is pixel-wise detection algorithm. We customized the architecture for semantic segmentation. We used YOLOv5s to train object detection model. After training, the inference time of semantic segmentation was compared to object detection. It was observed that object detection was slightly faster and semantic segmentation is slower but better at detecting the exact shape of the object. However, we found out that this slight difference of inference can grow exponentially if the both models get more complex. To solve this, we propose a novel approach called Adaptive distance detection which combines both algorithms. We will also introduce Adaptive instance segmentation to combine semantic segmentation and panoptic segmentation.

## How to Run Video Analytics for Manufacturing

DevFeed: [How to Run Video Analytics for Manufacturing](<https://devfeed.tech/articles/how-to-run-video-analytics-for-manufacturing-130063.md>)

Original publisher: [Read original article](<https://roboflow.com/blog/run-video-analytics-for-manufacturing>)

Author: Mostafa Ibrahim

Published: 2026-11-12T03:32:35Z

Content type: tutorial

Language: en

Sources: [Roboflow](<https://devfeed.tech/sources/roboflow-blog.md>)

Topics: [Video Analytics](<https://devfeed.tech/topics/video-analytics.md>), [image recognition](<https://devfeed.tech/topics/image-recognition.md>), [Local AI](<https://devfeed.tech/topics/local-ai.md>), [AI Infrastructure](<https://devfeed.tech/topics/ai-infrastructure.md>)

Tags: [alert](<https://devfeed.tech/tags/alert.md>), [alerts](<https://devfeed.tech/tags/alerts.md>), [analytics](<https://devfeed.tech/tags/analytics.md>), [api](<https://devfeed.tech/tags/api.md>), [api-key](<https://devfeed.tech/tags/api-key.md>), [architecture](<https://devfeed.tech/tags/architecture.md>), [bandwidth](<https://devfeed.tech/tags/bandwidth.md>), [bytetrack](<https://devfeed.tech/tags/bytetrack.md>), [camera](<https://devfeed.tech/tags/camera.md>), [cloud](<https://devfeed.tech/tags/cloud.md>), [computer-vision](<https://devfeed.tech/tags/computer-vision.md>), [deployment](<https://devfeed.tech/tags/deployment.md>), [edge](<https://devfeed.tech/tags/edge.md>), [factory](<https://devfeed.tech/tags/factory.md>), [guide](<https://devfeed.tech/tags/guide.md>), [how-to](<https://devfeed.tech/tags/how-to.md>), [manufacturing](<https://devfeed.tech/tags/manufacturing.md>), [roboflow-workflows](<https://devfeed.tech/tags/roboflow-workflows.md>), [video](<https://devfeed.tech/tags/video.md>), [video-analytics](<https://devfeed.tech/tags/video-analytics.md>), [workflows](<https://devfeed.tech/tags/workflows.md>)

### AI overview

Manufacturing video analytics can reuse RTSP cameras to detect visible defects, monitor safety zones, and count production, but results depend on camera views, model fit, and testing under factory conditions. In Roboflow Workflows, detection, tracking, and rule blocks support these tasks: ByteTrack assigns carton IDs and Line Counter counts crossings; inspection workflows can flag missing tablets or classify cropped tablets for cracks. The guide recommends starting with one measurable task, validating against manual counts or reviewed footage, and checking missed detections and false alerts. In its carton test, lowering the confidence threshold from 0.3 to 0.25 recovered a missed carton, with a stated risk of false detections. Self-hosted Inference processes streams on factory hardware and avoids credits but requires maintained, capable equipment. Standard Serverless API suits occasional image requests; Cloud Streaming handles managed live RTSP/ONVIF cameras; Dedicated Deployments provide private managed cloud resources. Cloud live streaming depends on reliable internet, while local processing can support immediate decisions.

### Source excerpt

A practical guide to using Roboflow Workflows with factory camera feeds and choosing between edge and cloud deployment.

## Pulse or Provenance? What rPPG-Based Deepfake Detectors Learn

DevFeed: [Pulse or Provenance? What rPPG-Based Deepfake Detectors Learn](<https://devfeed.tech/articles/pulse-or-provenance-what-rppg-based-deepfake-detectors-learn-128783.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.13992>)

Author: Othmane Harraq, Tamer Aldwairi

Published: 2026-11-11T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Cryptography and Security](<https://devfeed.tech/sources/arxiv-cryptography-and-security.md>), [arXiv: Machine Learning](<https://devfeed.tech/sources/arxiv-machine-learning.md>)

Topics: [Machine learning](<https://devfeed.tech/topics/machine-learning.md>), [Video Analytics](<https://devfeed.tech/topics/video-analytics.md>), [Image processing](<https://devfeed.tech/topics/image-processing.md>)

Tags: [background](<https://devfeed.tech/tags/background.md>), [band](<https://devfeed.tech/tags/band.md>), [clip](<https://devfeed.tech/tags/clip.md>), [cs-cr](<https://devfeed.tech/tags/cs-cr.md>), [cs-cv](<https://devfeed.tech/tags/cs-cv.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [deepfake](<https://devfeed.tech/tags/deepfake.md>), [deepfake-detection](<https://devfeed.tech/tags/deepfake-detection.md>), [detection](<https://devfeed.tech/tags/detection.md>), [encoding](<https://devfeed.tech/tags/encoding.md>), [experiments](<https://devfeed.tech/tags/experiments.md>), [generators](<https://devfeed.tech/tags/generators.md>), [manipulation](<https://devfeed.tech/tags/manipulation.md>), [order](<https://devfeed.tech/tags/order.md>), [protocols](<https://devfeed.tech/tags/protocols.md>), [regions](<https://devfeed.tech/tags/regions.md>), [signal](<https://devfeed.tech/tags/signal.md>), [simple](<https://devfeed.tech/tags/simple.md>), [synthetic](<https://devfeed.tech/tags/synthetic.md>), [test](<https://devfeed.tech/tags/test.md>), [traces](<https://devfeed.tech/tags/traces.md>), [training](<https://devfeed.tech/tags/training.md>), [video](<https://devfeed.tech/tags/video.md>)

### AI overview

The reported rPPG-style signal appears to track video provenance and generation artifacts rather than cardiac activity. On talking-face forgeries, simple chrominance traces achieve 0.998 AUC, above a dedicated rPPG extractor at 0.87; experiments locate the discriminative signal above the heart-rate band and show it in backgrounds and full frames. Removing high-frequency content makes most synthesized videos appear real, while adding it makes real videos appear fake. Face-based detectors trained on Celeb-DF++ label over 99% of real FF++ videos as fake, and a frozen CLIP backbone matches or exceeds feature fusions. The abstract also reports identity and source-video overlap in official Celeb-DF splits, with reuse in custom protocols inflating weaker detectors by up to 0.30 AUC.

### Source excerpt

arXiv:2611.13992v1 Announce Type: cross Abstract: Remote photoplethysmography (rPPG) has been proposed as a physiological cue for deepfake detection, based on the premise that synthesized faces lack a genuine cardiac signal. We examine what rPPG-style detectors learn on Celeb-DF++ (22 forgery methods) and FaceForensics++ (FF++), using one identically sampled window per real and fake video. On talking-face forgeries, simple chrominance traces reach an AUC of 0.998, exceeding a dedicated rPPG extractor (0.87). Controlled experiments indicate that this signal is not physiological in origin. On fully synthesized forgeries, it is equally discriminative when extracted from background regions or from full frames without face detection, and it is concentrated above the 0.7 to 4 Hz heart-rate band. It is largely removed by shuffling frame order but persists after normalizing resolution, frame rate and encoding, and an independent reimplementation reproduces these findings. Removing high-frequency content makes most synthesized videos appear real, adding it makes real videos appear fake, and injecting a synthetic pulse has little effect. In controlled self-reenactment experiments with two generators, animating a face from a single frame introduces the same high-frequency signal, whereas reconstructing each frame from its own source does not. Face-based chrominance detectors trained on Celeb-DF++ classify over 99% of real FF++ videos as fake, indicating that they largely encode video source rather than manipulation, and a CLIP-based detector exhibits the same behavior. A frozen CLIP backbone matches or exceeds fusions of CLIP with rPPG and frequency features. Finally, we show that the official Celeb-DF and Celeb-DF++ splits share identities and source-video content between training and test, and that custom protocols reusing test videos and identities in training inflate weaker detectors by up to 0.30 AUC.

## Obscured Wildfire Flame Detection by Spatio-temporal Analysis of Smoke Patterns Using Frame-wise Transformers

DevFeed: [Obscured Wildfire Flame Detection by Spatio-temporal Analysis of Smoke Patterns Using Frame-wise Transformers](<https://devfeed.tech/articles/obscured-wildfire-flame-detection-by-spatio-temporal-analysis-of-smoke-patterns-using-frame-wise-transformers-128393.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.14074>)

Author: Uma Meleti, Abolfazl Razi, Fatemeh Afghah

Published: 2026-11-11T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>)

Topics: [image recognition](<https://devfeed.tech/topics/image-recognition.md>), [Video Analytics](<https://devfeed.tech/topics/video-analytics.md>), [Transformers](<https://devfeed.tech/topics/transformers.md>), [Image processing](<https://devfeed.tech/topics/image-processing.md>), [Artificial Intelligence](<https://devfeed.tech/topics/ai.md>)

Tags: [ablation](<https://devfeed.tech/tags/ablation.md>), [architecture](<https://devfeed.tech/tags/architecture.md>), [attention](<https://devfeed.tech/tags/attention.md>), [classification](<https://devfeed.tech/tags/classification.md>), [classification-accuracy](<https://devfeed.tech/tags/classification-accuracy.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-cv](<https://devfeed.tech/tags/cs-cv.md>), [dataset](<https://devfeed.tech/tags/dataset.md>), [deep-learning](<https://devfeed.tech/tags/deep-learning.md>), [detection](<https://devfeed.tech/tags/detection.md>), [drones](<https://devfeed.tech/tags/drones.md>), [fire](<https://devfeed.tech/tags/fire.md>), [flame](<https://devfeed.tech/tags/flame.md>), [image-processing](<https://devfeed.tech/tags/image-processing.md>), [ir](<https://devfeed.tech/tags/ir.md>), [model](<https://devfeed.tech/tags/model.md>), [object-detection](<https://devfeed.tech/tags/object-detection.md>), [optimize](<https://devfeed.tech/tags/optimize.md>), [pattern](<https://devfeed.tech/tags/pattern.md>), [performance](<https://devfeed.tech/tags/performance.md>), [transformers](<https://devfeed.tech/tags/transformers.md>), [wildfire](<https://devfeed.tech/tags/wildfire.md>)

### AI overview

The method detects visible and obscured wildfire flames by analyzing smoke motion across consecutive RGB video frames. Its frame-wise transformer applies attention between frames while retaining within-frame spatial information, and adjusts temporal pattern lengths for different resolutions to balance performance and computational cost. On a curated FLAME2 dataset with dual-mode RGB/IR video, IR supplies ground truth while training uses RGB alone, supporting use with regular-camera drones and RGB-only datasets. Reported results are 92.61% foreground Dice, 93.21% detection precision, 91.73% recall, and 98.02% clip-level classification accuracy. The abstract also reports ablation study of efficacy and generalizability, but gives no further details.

### Source excerpt

arXiv:2611.14074v1 Announce Type: cross Abstract: In this paper, we propose a new paradigm in wildfire management for detecting both visible and obscured fire using Deep Learning (DL). The essence of our method is the joint spatial and temporal analysis of RGB videos, embracing the fact that temporal patterns of smoke motions in consecutive video frames can indicate obscured fire flames. This perspective contrasts the traditional approach of independently analyzing images and individual video frames. To this end, we propose a frame-wise transformer architecture that applies an attention mechanism between the consecutive frames while preserving spatial information within frames as opposed to regular attention mechanisms like the non-local block, which combines spatial and temporal information into a single vector. We also optimize the temporal pattern lengths under different spatial resolutions to achieve the highest performance while reducing the computational cost. To investigate the applicability of our method in aerial image processing, we applied it to a curated version of the FLAME2 dataset that includes dual-mode RGB/IR videos. The IR video feed is used to extract ground truth, while the model is trained only on RGB videos to maintain applicability to lowcost commercial drones (which are equipped only with regular cameras) and facilitate processing existing RGB-only datasets. Our proposed method achieves a Segmentation Foreground Dice score of 92.61%, an Object Detection Precision and Recall rate of 93.21% and 91.73%, and clip-level classification accuracy of 98.02%, respectively. We also investigated the efficacy and generalizability of our method through ablation study

## From Weapons to Events: Evaluating Evidential Support in Multimodal AI for Policing

DevFeed: [From Weapons to Events: Evaluating Evidential Support in Multimodal AI for Policing](<https://devfeed.tech/articles/from-weapons-to-events-evaluating-evidential-support-in-multimodal-ai-for-policing-125904.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.10245>)

Author: Armin Mustafa, Syed Sameed Husain

Published: 2026-11-10T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>)

Topics: [Computer vision](<https://devfeed.tech/topics/computer-vision.md>), [multimodal-ai](<https://devfeed.tech/topics/multimodal-ai.md>), [Video Analytics](<https://devfeed.tech/topics/video-analytics.md>), [image recognition](<https://devfeed.tech/topics/image-recognition.md>)

Tags: [ai](<https://devfeed.tech/tags/ai.md>), [computer-vision](<https://devfeed.tech/tags/computer-vision.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-cv](<https://devfeed.tech/tags/cs-cv.md>), [cs-cy](<https://devfeed.tech/tags/cs-cy.md>), [dataset](<https://devfeed.tech/tags/dataset.md>), [detection](<https://devfeed.tech/tags/detection.md>), [evaluation](<https://devfeed.tech/tags/evaluation.md>), [events](<https://devfeed.tech/tags/events.md>), [law-enforcement](<https://devfeed.tech/tags/law-enforcement.md>), [multimodal](<https://devfeed.tech/tags/multimodal.md>), [multimodal-ai](<https://devfeed.tech/tags/multimodal-ai.md>), [pipeline](<https://devfeed.tech/tags/pipeline.md>), [pipelines](<https://devfeed.tech/tags/pipelines.md>), [protocol](<https://devfeed.tech/tags/protocol.md>), [retrieval](<https://devfeed.tech/tags/retrieval.md>), [verification](<https://devfeed.tech/tags/verification.md>), [video](<https://devfeed.tech/tags/video.md>), [weapons](<https://devfeed.tech/tags/weapons.md>)

### AI overview

The paper evaluates policing AI at two levels. For knife retrieval, its UK law enforcement-developed KnifeHunter dataset contains over 25,000 images across 543 knife classes; KHNet, which fuses local and global features, achieves 85.7% mAP on the medium protocol and 63.4% on the hard protocol with about 125,000 distractor images. For body-worn video, the authors score complete pipelines on transcription, speaker attribution, event detection, timing, and unsupported events, using a judge validated against blinded human labels. On the 27-recording, 118-minute UKBWV-27 pilot stress test, their NarrativeBridge-PAL-GPT-5 hybrid leads on five of six measures, but 13.6% of its predicted events are unsupported and 27.7% of utterances are misattributed. All pipelines degrade on edited releases, long recordings, and noisy audio. The authors propose readiness levels that distinguish assistive use from tasks requiring verification of every output.

### Source excerpt

arXiv:2611.10245v1 Announce Type: cross Abstract: Computer vision (CV) for policing has centred on recognising objects such as seized weapons, while body-worn video (BWV) is still reviewed by people, with existing AI tools using only its audio. Multimodal video AI could make BWV review faster and cheaper. At the object level, we present the KnifeHunter dataset, developed with UK law enforcement, with over 25,000 images of 543 knife classes and a protocol reflecting investigative conditions, and KHNet, a retrieval architecture fusing local and global features, which with about 125,000 distractor images attains 85.7% and 63.4% mean Average Precision (mAP) on the medium and hard protocols. At the event level, we present an evaluation protocol for video understanding pipelines, which describe a video as events and a transcript, that scores transcription, speaker attribution, event detection, temporal localisation and unsupported events. Whether each event is supported by the recording is labelled by a judge validated against blinded human labels. We apply it to complete pipelines, three 2026 omni-modal models and our hybrid NarrativeBridge-PAL-GPT-5 pipeline, on UKBWV-27, a pilot stress-test set of 27 publicly released UK BWV recordings (118 minutes). On this set the hybrid pipeline leads on five of six measures, yet it leaves 13.6% of predicted events unsupported and misattributes 27.7% of utterances, and every pipeline degrades on recordings from edited releases, long recordings and noisy audio. We propose readiness levels separating assistive tasks from those requiring verification of every output.

## Beyond Hash Matching: AI-Based Detection of Previously Unidentified CSAM and its Legal Implications in the EU

DevFeed: [Beyond Hash Matching: AI-Based Detection of Previously Unidentified CSAM and its Legal Implications in the EU](<https://devfeed.tech/articles/beyond-hash-matching-ai-based-detection-of-previously-unidentified-csam-and-its-legal-implications-in-the-eu-125864.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.10028>)

Author: Emmanouela Kokolaki, Georgios Samaritakis, Paraskevi Fragopoulou

Published: 2026-11-10T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>)

Topics: [Video Analytics](<https://devfeed.tech/topics/video-analytics.md>), [ai regulation](<https://devfeed.tech/topics/ai-regulation.md>), [AI Integration Strategies](<https://devfeed.tech/topics/ai-integration-strategies.md>), [human review](<https://devfeed.tech/topics/human-review.md>), [image tagging](<https://devfeed.tech/topics/image-tagging.md>), [sensitive data](<https://devfeed.tech/topics/sensitive-data.md>)

Tags: [ai](<https://devfeed.tech/tags/ai.md>), [classification](<https://devfeed.tech/tags/classification.md>), [computer-vision](<https://devfeed.tech/tags/computer-vision.md>), [content-moderation](<https://devfeed.tech/tags/content-moderation.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-cv](<https://devfeed.tech/tags/cs-cv.md>), [csam](<https://devfeed.tech/tags/csam.md>), [data-protection](<https://devfeed.tech/tags/data-protection.md>), [detection](<https://devfeed.tech/tags/detection.md>), [eu](<https://devfeed.tech/tags/eu.md>), [eu-ai-act](<https://devfeed.tech/tags/eu-ai-act.md>), [evaluation](<https://devfeed.tech/tags/evaluation.md>), [hash](<https://devfeed.tech/tags/hash.md>), [human-oversight](<https://devfeed.tech/tags/human-oversight.md>), [integration](<https://devfeed.tech/tags/integration.md>), [legal](<https://devfeed.tech/tags/legal.md>), [prioritization](<https://devfeed.tech/tags/prioritization.md>), [real-time](<https://devfeed.tech/tags/real-time.md>), [regulation](<https://devfeed.tech/tags/regulation.md>), [reporting](<https://devfeed.tech/tags/reporting.md>), [transparency](<https://devfeed.tech/tags/transparency.md>), [video](<https://devfeed.tech/tags/video.md>)

### AI overview

Traditional hash matching mainly finds previously verified CSAM, so the paper reviews AI approaches that could help detect, analyze, and prioritize previously unidentified material. It groups them into real-time image and video analysis, content classification and tagging, and behavioral monitoring, and compares general-purpose vision platforms, moderation services, and specialized detection tools. An evaluation of Google Cloud Vision AI and Sightengine found differences in classification outputs and significant limits in estimating age, especially for adolescents. The authors therefore frame AI assessments as aids to analysis and prioritization, not definitive CSAM determinations. Their EU legal review highlights how obligations under the AI Act and GDPR, and safeguards for rights, transparency, accountability, and human oversight, may vary by deployment context and lead to inconsistencies across detection and reporting workflows.

### Source excerpt

arXiv:2611.10028v1 Announce Type: cross Abstract: The detection of previously unidentified Child Sexual Abuse Material (CSAM) remains a significant challenge for law enforcement authorities, online service providers, and child protection organizations, as traditional hash-matching technologies are primarily designed to identify previously verified material. This paper reviews and classifies existing AI-based technologies that could complement hash-based approaches and support the detection, analysis, and prioritization of potentially harmful content. The reviewed approaches are organized into three categories: real-time image and video analysis, image and video classification and tagging, and AI-based behavioral monitoring. A comparative analysis examines the capabilities, limitations, and potential integration of general-purpose computer vision platforms, content moderation services, and specialized CSAM detection tools within existing detection workflows. An experimental evaluation was conducted using Google Cloud Vision AI and Sightengine to investigate sexual-content classification and child-adult age estimation. The results highlight differences in classification taxonomies and outputs across providers, as well as significant limitations in age classification, particularly for adolescents. These findings highlight the importance of using AI-based assessments as complementary tools for analysis and prioritization, rather than as definitive determination of CSAM. Finally, the paper examines the relevant European Union regulatory framework, with particular attention to the EU AI Act, the General Data Protection Regulation (GDPR), fundamental rights, transparency, accountability, and human oversight. The legal analysis highlights how regulatory obligations and safeguards may differ depending on the deployment context, potentially creating inconsistencies across interconnected detection and reporting workflows.

## MARS: Multiscale Appearance-Residual Streams for Video Forgery Detection

DevFeed: [MARS: Multiscale Appearance-Residual Streams for Video Forgery Detection](<https://devfeed.tech/articles/mars-multiscale-appearance-residual-streams-for-video-forgery-detection-126285.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.12508>)

Author: Ahmed Y. Radwan, Shaina Raza

Published: 2026-11-10T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>)

Topics: [Video Analytics](<https://devfeed.tech/topics/video-analytics.md>)

Tags: [appearance](<https://devfeed.tech/tags/appearance.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-cv](<https://devfeed.tech/tags/cs-cv.md>), [datasets](<https://devfeed.tech/tags/datasets.md>), [deepfake](<https://devfeed.tech/tags/deepfake.md>), [deepfake-detection](<https://devfeed.tech/tags/deepfake-detection.md>), [detection](<https://devfeed.tech/tags/detection.md>), [difference](<https://devfeed.tech/tags/difference.md>), [features](<https://devfeed.tech/tags/features.md>), [forgery](<https://devfeed.tech/tags/forgery.md>), [frame](<https://devfeed.tech/tags/frame.md>), [gate](<https://devfeed.tech/tags/gate.md>), [intermediate](<https://devfeed.tech/tags/intermediate.md>), [manipulation](<https://devfeed.tech/tags/manipulation.md>), [mars](<https://devfeed.tech/tags/mars.md>), [motion](<https://devfeed.tech/tags/motion.md>), [operators](<https://devfeed.tech/tags/operators.md>), [patch](<https://devfeed.tech/tags/patch.md>), [predictions](<https://devfeed.tech/tags/predictions.md>), [protocol](<https://devfeed.tech/tags/protocol.md>), [streams](<https://devfeed.tech/tags/streams.md>), [video](<https://devfeed.tech/tags/video.md>)

### AI overview

MARS combines two video-forgery detection paths: one applies fixed spatial and temporal difference-of-Gaussians operators to frozen intermediate features and models residuals with appearance context; the other adapts encoder normalization parameters for frame classification. A clip-dependent gate combines their predictions. Across seven datasets under a shared protocol, it reports the highest average AUROC among seven compared detectors (91.43%). Trained only on facial manipulations, it reports 93.99% AUROC on GenVideo and 75.50% on DF26, indicating transfer to generated videos without generated-video training data. It also provides signed patch and frame contributions, evaluated by removing spatial and temporal regions and comparing results with random controls.

### Source excerpt

arXiv:2611.12508v1 Announce Type: cross Abstract: Generalizing deepfake detection across facial manipulation and fully generated video remains challenging due to forgery traces varying with the manipulation method and scene content. Explicit spatial and temporal feature differences offer a way to capture this variation, but also respond to natural texture and motion, motivating their joint interpretation with appearance features. We investigate whether combining multiscale feature differences with adapted appearance representations improves detection across both forgery categories. We introduce Multiscale Appearance-Residual Streams (MARS), a two-path video detector. The residual path applies fixed spatial and temporal difference-of-Gaussians operators to frozen intermediate features and models the resulting differences alongside appearance context across frames. The semantic path adapts the encoder's normalization parameters for frame classification, and a clip-dependent gate combines their predictions. Under a shared protocol across seven datasets, MARS achieves the highest average AUROC of 91.43\%, outperforming seven compared detectors. When trained on facial manipulations alone, MARS achieves 93.99\% AUROC on GenVideo and 75.50\% on DF26, demonstrating transfer to generated videos without generated-video training data. MARS also provides signed patch and frame contributions, whose faithfulness we evaluate through controlled spatial and temporal removal against random controls. MARS is publicly available for research: Project page: https://vectorinstitute.github.io/mars/ Github: https://github.com/VectorInstitute/mars

## Counting Events, Not Seeing Them: Diagnosing and Repairing the Readout Failure in Video LLMs

DevFeed: [Counting Events, Not Seeing Them: Diagnosing and Repairing the Readout Failure in Video LLMs](<https://devfeed.tech/articles/counting-events-not-seeing-them-diagnosing-and-repairing-the-readout-failure-in-video-llms-125113.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.09949>)

Author: Dongwu Qian

Published: 2026-11-10T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Machine Learning](<https://devfeed.tech/sources/arxiv-machine-learning.md>)

Topics: [Video Analytics](<https://devfeed.tech/topics/video-analytics.md>), [Machine learning](<https://devfeed.tech/topics/machine-learning.md>), [Decoding](<https://devfeed.tech/topics/decoding.md>)

Tags: [2](<https://devfeed.tech/tags/2.md>), [3](<https://devfeed.tech/tags/3.md>), [abstract](<https://devfeed.tech/tags/abstract.md>), [arxiv](<https://devfeed.tech/tags/arxiv.md>), [attribution](<https://devfeed.tech/tags/attribution.md>), [audit](<https://devfeed.tech/tags/audit.md>), [bench](<https://devfeed.tech/tags/bench.md>), [benchmark](<https://devfeed.tech/tags/benchmark.md>), [best](<https://devfeed.tech/tags/best.md>), [boundary](<https://devfeed.tech/tags/boundary.md>), [budget](<https://devfeed.tech/tags/budget.md>), [but](<https://devfeed.tech/tags/but.md>), [capacity](<https://devfeed.tech/tags/capacity.md>), [channel](<https://devfeed.tech/tags/channel.md>), [clip](<https://devfeed.tech/tags/clip.md>), [concurrent](<https://devfeed.tech/tags/concurrent.md>), [cs-cl](<https://devfeed.tech/tags/cs-cl.md>), [cs-cv](<https://devfeed.tech/tags/cs-cv.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [events](<https://devfeed.tech/tags/events.md>), [failure](<https://devfeed.tech/tags/failure.md>), [llms](<https://devfeed.tech/tags/llms.md>), [video](<https://devfeed.tech/tags/video.md>)

### AI overview

In synthetic tests, video LLMs encode object presence within individual windows, but global event counts are poorly read out and confounded with video duration. Episode-localization supervision and a discrete event-symbol side channel raise same-object counting from 46.7% to 65.0% on a chance-33% benchmark, near a 66.7% oracle. On REALCOUNT, the best unaided result is 55.6% at 33.3% chance, and symbol transfer fails zero-shot. Longer adaptation improves results: the symbol-channel reader reaches 94.4% with sequence-disjoint oracle timelines after three epochs, and 72.2% with deployed-detector timelines after two. The remaining limitation appears tied to weak, poorly transferring detector evidence: a boundary probe reports AUC 0.62 versus a 0.51 permutation control.

### Source excerpt

arXiv:2611.09949v1 Announce Type: cross Abstract: Video LLMs fail conspicuously at counting how many times an object crosses the screen, a failure attributed to missing object permanence. Using construction-grounded synthetic testbeds with independently validated labels, we show this attribution is incomplete: the count signal is present but unread - linear probes decode per-window presence at 93% while global count decoding collapses to a duration confound. Concurrent work reports the same representation-output dissociation for repetition counting; we show that in exit-and-reappear event counting it is repairable: two readout-directed repairs (episode-localization supervision and a discrete event-symbol side channel) raise same-object counting from 46.7% to 65.0% (oracle 66.7%) on a chance-33% benchmark across 7B-30B systems. Transfer to an independent factory costs nothing, and an audit of the measurement apparatus exposes four data-integrity failures that had obscured the boundary. A real-video audit (REALCOUNT, n=18, blind-audited) shows the boundary persists - best 55.6% at chance 33.3% - and symbol transfer fails zero-shot. A 1-epoch in-domain adaptation bench is negative (six levers at 28-50%), but a 3-epoch schedule shows this is budget-limited for the symbol channel: with a 3-epoch schedule and sequence-disjoint oracle timelines the reader reaches 94.4% on REALCOUNT, and with the deployed detector's timelines a 2-epoch run reaches 72.2%. The residual gap is detector evidence and the reader's match to detector noise, not reader capacity: a frozen-feature boundary probe finds the signal weak (AUC 0.62 over a 0.51 permutation control) and non-transferring across clips (0.77 within-clip vs. 0.62 across windows).

## Beyond Frame-Level Accuracy: Analyzing Video-Level Stability in Transformer-Based Deepfake Forensics

DevFeed: [Beyond Frame-Level Accuracy: Analyzing Video-Level Stability in Transformer-Based Deepfake Forensics](<https://devfeed.tech/articles/beyond-frame-level-accuracy-analyzing-video-level-stability-in-transformer-based-deepfake-forensics-125156.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.10249>)

Author: Syed Azmul Hasan, Mowmita Pervin Hera, Afroza Lona Mim, M M Jihad Hossain, Hazrat Ali, Mohammad Farhad Bulbul

Published: 2026-11-10T05:00:00Z

Content type: comparison

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>), [arXiv: Machine Learning](<https://devfeed.tech/sources/arxiv-machine-learning.md>)

Topics: [Transformer](<https://devfeed.tech/topics/transformer.md>), [Machine learning](<https://devfeed.tech/topics/machine-learning.md>), [Video Analytics](<https://devfeed.tech/topics/video-analytics.md>), [reproducibility](<https://devfeed.tech/topics/reproducibility.md>)

Tags: [attention](<https://devfeed.tech/tags/attention.md>), [cnn](<https://devfeed.tech/tags/cnn.md>), [convolutional-neural-networks](<https://devfeed.tech/tags/convolutional-neural-networks.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-cv](<https://devfeed.tech/tags/cs-cv.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [dataset](<https://devfeed.tech/tags/dataset.md>), [deepfake](<https://devfeed.tech/tags/deepfake.md>), [detection](<https://devfeed.tech/tags/detection.md>), [forensics](<https://devfeed.tech/tags/forensics.md>), [frame](<https://devfeed.tech/tags/frame.md>), [performance](<https://devfeed.tech/tags/performance.md>), [transformer](<https://devfeed.tech/tags/transformer.md>), [video](<https://devfeed.tech/tags/video.md>)

### AI overview

In a controlled comparison using the same face-centric pipeline and data splits, ViT-B/16 showed more stable video-level deepfake detection than ResNet-18 and EfficientNet-B0, whose frame-level performance weakened after video aggregation. Across three seeds, ViT-B/16 reached 92.8% video-level accuracy and 0.950 AUC on Celeb-DF v2. Cross-dataset AUC was asymmetric (0.79 from FaceForensics++ to Celeb-DF v2, versus 0.64 in the reverse direction), so in-domain stability did not carry over reliably to unseen manipulation domains.

### Source excerpt

arXiv:2611.10249v1 Announce Type: cross Abstract: Deepfake videos generated by advanced generative models pose serious threats to digital security, misinformation control, and media forensics. Although numerous detection methods have been proposed, their reported performances are often based on heterogeneous preprocessing pipelines, inconsistent data splits, and model-specific training protocols, hindering fair architectural comparison. This work presents a controlled and reproducible comparative study of convolutional neural networks (ResNet-18 and EfficientNet-B0) and a transformer-based architecture (ViT-B/16) within a unified, face-centric deepfake detection framework. All models are trained using identical preprocessing, augmentation, optimization settings, and data splits on Celeb-DF v2 and FaceForensics++, and results are validated across three independent random seeds with bootstrapped confidence intervals. Experimental results show that while CNN-based models perform competitively at the frame level, their performance degrades after video-level aggregation. In contrast, ViT-B/16 demonstrates more stable video-level performance, achieving 92.8\% [95\% CI: 91.7--94.0] video-level accuracy and an AUC of 0.950 [95\% CI: 0.934--0.964] on Celeb-DF v2 (3-seed bootstrapped). Cross-dataset evaluation across the same three random seeds reveals a large, asymmetric generalization gap (AUC 0.79 FF++$\rightarrow$Celeb-DF vs. 0.64 Celeb-DF$\rightarrow$FF++), indicating that in-domain video-level stability does not transfer to unseen manipulation domains. These findings indicate that global self-attention improves robustness and video-level stability for deepfake detection within a given data domain, while highlighting the need for further work on cross-domain generalization.

## Multimodal Training with Knowledge Traction for Video Anomaly Detection

DevFeed: [Multimodal Training with Knowledge Traction for Video Anomaly Detection](<https://devfeed.tech/articles/multimodal-training-with-knowledge-traction-for-video-anomaly-detection-124698.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.09536>)

Author: Xiao Tan

Published: 2026-11-09T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>)

Topics: [Video Analytics](<https://devfeed.tech/topics/video-analytics.md>), [multimodal-ai](<https://devfeed.tech/topics/multimodal-ai.md>), [Representation learning](<https://devfeed.tech/topics/representation-learning.md>), [Large Language Model](<https://devfeed.tech/topics/llm.md>), [Model Development](<https://devfeed.tech/topics/model-development.md>)

Tags: [anomaly-detection](<https://devfeed.tech/tags/anomaly-detection.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-cv](<https://devfeed.tech/tags/cs-cv.md>), [knowledge](<https://devfeed.tech/tags/knowledge.md>), [large-language-model](<https://devfeed.tech/tags/large-language-model.md>), [multimodal](<https://devfeed.tech/tags/multimodal.md>), [training](<https://devfeed.tech/tags/training.md>), [video](<https://devfeed.tech/tags/video.md>)

### AI overview

The method adds external knowledge to weakly supervised video anomaly detection, where existing approaches largely judge anomalies from video-frame feature magnitudes. It uses textual prompts and an LLM to establish task-specific priors, then dictionary learning to normalize and compress them. To address inconsistent captions for similar frames, it uses an LLM and auxiliary information to make descriptions more consistent within a video. Prior knowledge and cross-modal attention are then used to learn relationships among modalities and build feature representations. The abstract reports improved accuracy and stability and says the method surpasses other state-of-the-art approaches, but gives no metrics or experimental details.

### Source excerpt

arXiv:2611.09536v1 Announce Type: cross Abstract: The task of weakly supervised video anomaly detection is challenging. Existing approaches mainly model video frame features and determine whether anomalies are present based on their magnitudes. In this work, we propose a novel method that incorporates external knowledge into training to improve model performance, robustness, and generalization. We leverage textual prompts and a large language model (LLM) to establish prior information for the specific task, and employ dictionary learning to normalize and compress this information. Moreover, during experiments, we observed inconsistencies in the textual descriptions extracted from similar video frames by image caption models. To address this, we devise a method that leverages an LLM and auxiliary information to ensure consistency in the textual descriptions of similar frames within a video. Finally, we introduce a feature processing approach that incorporates prior knowledge and cross-modal attention mechanisms. This approach learns the relationships and characteristics among different modalities to construct more accurate and reliable feature representations, thereby improving the accuracy and stability of downstream tasks. Experimental results demonstrate that our method surpasses other state-of-the-art approaches.

## Beyond Risk Cues: A Mechanism-Level Benchmark and Verifiable Debate for Scam Video Detection on the Web

DevFeed: [Beyond Risk Cues: A Mechanism-Level Benchmark and Verifiable Debate for Scam Video Detection on the Web](<https://devfeed.tech/articles/beyond-risk-cues-a-mechanism-level-benchmark-and-verifiable-debate-for-scam-video-detection-on-the-web-118423.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.06110>)

Author: Yifan Wang, Junyu Lu, Shun Zhang, Chaozhuo Li, Fanliang Bu, Rong Yin, Lingbin Bu, Changrong Min, Peng Chen

Published: 2026-11-05T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>)

Topics: [Video Analytics](<https://devfeed.tech/topics/video-analytics.md>), [AI Foundation Models](<https://devfeed.tech/topics/ai-foundation-models.md>), [generative-ai-tools](<https://devfeed.tech/topics/generative-ai-tools.md>)

Tags: [benchmark](<https://devfeed.tech/tags/benchmark.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [dataset](<https://devfeed.tech/tags/dataset.md>), [detection](<https://devfeed.tech/tags/detection.md>), [framework](<https://devfeed.tech/tags/framework.md>), [mechanism](<https://devfeed.tech/tags/mechanism.md>), [scam](<https://devfeed.tech/tags/scam.md>), [video](<https://devfeed.tech/tags/video.md>)

### AI overview

MechScam is a 6,614-video benchmark designed to assess scam mechanisms, contextualized video evidence, and confusion with similar non-scam content such as advertisements. It includes evidence-based explanations, annotations of potential scam elements, and misleading arguments built from genuine evidence. ScamCourt has trained prosecution and defense debaters present opposing arguments, then a trained judge decide from a transcript with verified citations. The abstract reports that state-of-the-art multimodal language models struggle with similar non-scam content; ScamCourt outperforms larger models and fine-tuned baselines using the same backbone, with fewer false positives and greater resistance to misleading arguments.

### Source excerpt

arXiv:2611.06110v1 Announce Type: new Abstract: Online scam videos pose significant threats to individuals and society. Existing studies primarily focus on binary classification, offering limited insights into how scams work and the reasoning behind detection decisions. Moreover, they often overlook the challenge of distinguishing scam videos from superficially similar non-scam content, such as advertisements, and the robustness of detection models against misleading arguments. These limitations hinder reliable scam video detection. To address these limitations, we introduce MechScam, a benchmark comprising 6,614 videos covering diverse scam mechanisms and non-scam categories. MechScam provides explanations supported by video evidence, annotations of potential scam elements and their context, and misleading arguments constructed from genuine evidence. Recognizing that genuine evidence can still be used to justify misleading conclusions, we further propose ScamCourt, a debate framework in which trained prosecution and defense debaters present opposing arguments, while a trained judge makes decisions based on a transcript with verified citations. Extensive experiments show that state-of-the-art MLLMs struggle to distinguish scam videos from similar non-scam content. ScamCourt outperforms larger MLLMs and same-backbone fine-tuned baselines, reducing false positives and showing stronger resistance to misleading arguments. Together, MechScam and ScamCourt contribute to more reliable and interpretable scam video detection. The dataset and code are available at https://github.com/SuperYFan/MechScam.

[Next page](<https://devfeed.tech/topics/video-analytics.md?cursor=WyIyMDI2LTExLTA1VDA1OjAwOjAwKzAwOjAwIiwgIjY2ZTkxNWY2LWQxN2UtNDVmOC05M2I4LWQ5NWNmY2JlYjk0MiJd>)