# fourier transform

A mathematical transform that maps a signal from the time domain to the frequency domain, representing the amount and phase of each frequency component.

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

## Quantum Nondecimated Wavelet Transform: Theory, Circuits, and Applications

DevFeed: [Quantum Nondecimated Wavelet Transform: Theory, Circuits, and Applications](<https://devfeed.tech/articles/quantum-nondecimated-wavelet-transform-theory-circuits-and-applications-144868.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2512.21478>)

Author: Brani Vidakovic

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

Content type: article

Language: en

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

Topics: [Quantum Computing](<https://devfeed.tech/topics/quantum-computing.md>), [fourier transform](<https://devfeed.tech/topics/fourier-transform.md>), [fourier analysis](<https://devfeed.tech/topics/fourier-analysis.md>)

Tags: [applications](<https://devfeed.tech/tags/applications.md>), [circuits](<https://devfeed.tech/tags/circuits.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [denoising](<https://devfeed.tech/tags/denoising.md>), [filter](<https://devfeed.tech/tags/filter.md>), [linear-combination](<https://devfeed.tech/tags/linear-combination.md>), [measurement](<https://devfeed.tech/tags/measurement.md>), [model](<https://devfeed.tech/tags/model.md>), [physics-data-an](<https://devfeed.tech/tags/physics-data-an.md>), [quant-ph](<https://devfeed.tech/tags/quant-ph.md>), [quantum](<https://devfeed.tech/tags/quantum.md>), [scaling](<https://devfeed.tech/tags/scaling.md>), [simulations](<https://devfeed.tech/tags/simulations.md>), [stat-co](<https://devfeed.tech/tags/stat-co.md>), [theory](<https://devfeed.tech/tags/theory.md>)

### AI overview

The transform applies circular shifts to an amplitude-encoded signal before one orthogonal wavelet transform. Tracing out the shift register yields level probabilities matching normalized nondecimated wavelet energies, invariant to circular shifts; the diagonal gives moving averages of squared coefficients. The authors derive a multinomial model for level counts and a way to read lag-one autocovariance for all but the coarsest detail level, while noting that direct Fourier-basis measurement is at least as precise and uses a simpler circuit for these global statistics. For denoising, block-encoded shrinkage is placed between the transform and its inverse; postselection implements cycle spinning exactly as a linear combination of unitaries. Shared level weights yield a zero-phase translation-equivariant filter, while local weights preserve equivariance when indexed by nondecimated translate. Examples include James Stein shrinkage, local empirical Wiener filtering, and turbulent wind scaling; gate simulations verify the construction, with costs reported separately for state preparation, weight estimation, and output.

### Source excerpt

arXiv:2512.21478v2 Announce Type: replace-cross Abstract: The quantum nondecimated wavelet transform uses a shift register to apply circular shifts to an amplitude encoded signal before a single orthogonal wavelet transform. We show that tracing out the shift register gives level probabilities equal to the normalized nondecimated wavelet energies, invariant under circular shifts. Its diagonal gives moving averages of squared coefficients. We derive a multinomial model for the level counts and show how to read the autocovariance at lag one for all but the coarsest detail level. For these global statistics, direct Fourier basis measurement of the input is at least as precise and uses a simpler circuit. For denoising, we insert a block encoded shrinkage operator between the transform and its inverse. Postselection on the shift register implements the cycle spinning average exactly as a linear combination of unitaries. Weights shared within each level give a translation equivariant filter with zero phase. Local weights also preserve translation equivariance when indexed by the nondecimated translate. The circuit provides access to individual decimated branches and combines them coherently. We illustrate the method with James Stein shrinkage, local empirical Wiener filtering, and a scaling analysis of turbulent wind velocity. Gate simulations verify the construction, and we report state preparation, weight estimation, and output costs separately.

## Asymmetric Optical Image Watermarking

DevFeed: [Asymmetric Optical Image Watermarking](<https://devfeed.tech/articles/asymmetric-optical-image-watermarking-143120.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.23350>)

Author: Guang Yang

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

Content type: article

Language: en

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

Topics: [Cryptography](<https://devfeed.tech/topics/cryptography.md>), [Image processing](<https://devfeed.tech/topics/image-processing.md>), [fourier transform](<https://devfeed.tech/topics/fourier-transform.md>), [ed25519](<https://devfeed.tech/topics/ed25519.md>), [public key](<https://devfeed.tech/topics/public-key.md>), [Image](<https://devfeed.tech/topics/image.md>)

Tags: [cs-cr](<https://devfeed.tech/tags/cs-cr.md>), [cs-cv](<https://devfeed.tech/tags/cs-cv.md>), [ed25519](<https://devfeed.tech/tags/ed25519.md>), [image](<https://devfeed.tech/tags/image.md>), [optical](<https://devfeed.tech/tags/optical.md>), [security](<https://devfeed.tech/tags/security.md>), [simulation](<https://devfeed.tech/tags/simulation.md>), [watermarking](<https://devfeed.tech/tags/watermarking.md>)

### AI overview

The proposed watermark combines publicly verifiable authorship with a secret-key optical carrier. Its first layer binds an Ed25519 signature to image content in Fourier magnitudes, allowing offline verification with a public key. The second uses a secret phase hologram to write a carrier detected by a matched-filter correlator. In scalar-diffraction simulations, the optical signer closely reproduced the digital pipeline and verified at least 99.5% of 877 images in each of eight post-sensor settings, while accepting none of 40,670 unmarked images. The carrier was detected on 99.8% of unattacked images with the matched key, but detection fell to zero with a 1% scale error or a one-bin mask shift. These are simulation results; no hardware was built. The authors distinguish security claims based on signature and content binding from measured optical performance.

### Source excerpt

arXiv:2611.23350v1 Announce Type: new Abstract: Watermark detectors held by a provider cannot be checked by third parties, and a party that can detect a mark can usually also forge it. We study a two-layer image watermark in which the Fourier transform at the core of the pipeline is performed by a coherent optical 4F unit. Layer 1 carries an Ed25519 signature bound to the image content in Fourier magnitudes, so anyone with the public key verifies offline. Layer 2 is a key-defined carrier written by a secret phase hologram and read by a matched-filter correlator. The optics is a scalar-diffraction simulation with finite modulator bits, pixel aperture, pupil, defocus, misalignment, and sensor noise. No hardware is built. All criteria were fixed in a hashed specification before any optical result. The simulated optical signer reproduces the digital pipeline (field error 1.7e-15, bit agreement 99.997%), verifies at least 99.5% of 877 images under each of 8 post-sensor settings, one unattacked, and accepts none of 40,670 unmarked images. The optical carrier is key-selective: the matched key detects it on 99.8% of the unattacked images, and detection falls to 0% under a scale error of 1% and to 0% under a mask shift of 1 bin. Security rests on the signature and the content binding, for which we separate what is provable from what is measured.

## Brownian Bridge Diffusion-Assisted Joint Channel Estimation and Data Detection for Jamming-Resilient OFDM Receivers

DevFeed: [Brownian Bridge Diffusion-Assisted Joint Channel Estimation and Data Detection for Jamming-Resilient OFDM Receivers](<https://devfeed.tech/articles/brownian-bridge-diffusion-assisted-joint-channel-estimation-and-data-detection-for-jamming-resilient-ofdm-receivers-142536.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2606.28778>)

Author: Honghan She, Yufan Cheng, Siya Huang, Pengyu Wang, Tieming Sun, Kaikai Yang

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

Content type: article

Language: en

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

Topics: [communications](<https://devfeed.tech/topics/communications.md>), [fourier analysis](<https://devfeed.tech/topics/fourier-analysis.md>), [fourier transform](<https://devfeed.tech/topics/fourier-transform.md>)

Tags: [algorithm](<https://devfeed.tech/tags/algorithm.md>), [bits](<https://devfeed.tech/tags/bits.md>), [bottleneck](<https://devfeed.tech/tags/bottleneck.md>), [bridge](<https://devfeed.tech/tags/bridge.md>), [challenges](<https://devfeed.tech/tags/challenges.md>), [channel](<https://devfeed.tech/tags/channel.md>), [communication](<https://devfeed.tech/tags/communication.md>), [complexity](<https://devfeed.tech/tags/complexity.md>), [components](<https://devfeed.tech/tags/components.md>), [contamination](<https://devfeed.tech/tags/contamination.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-it](<https://devfeed.tech/tags/cs-it.md>), [data](<https://devfeed.tech/tags/data.md>), [design](<https://devfeed.tech/tags/design.md>), [detection](<https://devfeed.tech/tags/detection.md>), [devices](<https://devfeed.tech/tags/devices.md>), [diffusion](<https://devfeed.tech/tags/diffusion.md>), [domain](<https://devfeed.tech/tags/domain.md>), [eess-sp](<https://devfeed.tech/tags/eess-sp.md>), [environment](<https://devfeed.tech/tags/environment.md>), [fourier-transform](<https://devfeed.tech/tags/fourier-transform.md>), [framework](<https://devfeed.tech/tags/framework.md>), [graph](<https://devfeed.tech/tags/graph.md>), [math-it](<https://devfeed.tech/tags/math-it.md>)

### AI overview

BBD-JCED addresses OFDM jamming that contaminates pilots and data across time and frequency. Its first module detects jammed bins in the short-time Fourier transform domain and suppresses dominant components with spectral notching. A Brownian bridge probabilistic recovery module then works from the suppressed observation toward coded bits to jointly estimate the channel and detect data. The authors use a variational lower bound to form a tractable training objective and train the two modules in separate stages. Simulations with representative jamming types report better suppression and bit recovery than the considered baselines, with competitive inference complexity; the excerpt gives no quantitative results or baseline details.

### Source excerpt

arXiv:2606.28778v2 Announce Type: replace-cross Abstract: In next-generation wireless networks, the growing density of devices and limited spectrum resources pose severe jamming challenges to fragile legitimate communication links in the wireless electromagnetic environment. Crucially, when jamming overlaps with pilot and data symbols in both time and frequency domains, it inflicts a severe bottleneck on receiver-side joint estimation and detection. Existing schemes often lack an effective framework to combat such jamming contamination, thereby failing to guarantee reliable transmission. To address this issue, we propose a Brownian bridge diffusion-assisted joint channel estimation and data detection (BBD-JCED) framework for jamming-resilient orthogonal frequency-division multiplexing (OFDM) receivers. The proposed framework consists of a jamming suppression module and a Brownian bridge-joint recovery module. The former identifies jamming-contaminated bins in the short-time Fourier transform (STFT) domain and suppresses the dominant jamming components through spectral notching. The latter establishes a Brownian bridge-based probabilistic graph to model the probabilistic transition process from the suppressed OFDM observation to the coded bits, and performs a reverse recovery process to enable enhanced joint channel estimation and data detection. Moreover, we derive a variational lower bound (VLB) to reformulate the intractable bit estimate problem into a tractable training objective, providing a probabilistic foundation for the loss function design. Furthermore, we develop a two-stage training algorithm to separately optimize the jamming suppression and Brownian bridge-joint recovery modules. Simulation results under representative jamming types demonstrate that the proposed framework achieves superior jamming suppression and bit recovery performance over the considered baseline schemes while maintaining competitive inference complexity.

## Sparse Integer Fourier Transform and Its Application In Machine Learning

DevFeed: [Sparse Integer Fourier Transform and Its Application In Machine Learning](<https://devfeed.tech/articles/sparse-integer-fourier-transform-and-its-application-in-machine-learning-141713.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.21976>)

Author: Jinyang Li, Jie Chen

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

Content type: article

Language: en

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

Topics: [fourier transform](<https://devfeed.tech/topics/fourier-transform.md>), [fourier analysis](<https://devfeed.tech/topics/fourier-analysis.md>), [Machine learning](<https://devfeed.tech/topics/machine-learning.md>), [Feature Engineering](<https://devfeed.tech/topics/feature-engineering.md>)

Tags: [arithmetic](<https://devfeed.tech/tags/arithmetic.md>), [big-data](<https://devfeed.tech/tags/big-data.md>), [classifier](<https://devfeed.tech/tags/classifier.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [cs-ms](<https://devfeed.tech/tags/cs-ms.md>), [dft](<https://devfeed.tech/tags/dft.md>), [diagnosis](<https://devfeed.tech/tags/diagnosis.md>), [digital-signal-processing](<https://devfeed.tech/tags/digital-signal-processing.md>), [engineering](<https://devfeed.tech/tags/engineering.md>), [features](<https://devfeed.tech/tags/features.md>), [fft](<https://devfeed.tech/tags/fft.md>), [fourier-transform](<https://devfeed.tech/tags/fourier-transform.md>), [machine-learning](<https://devfeed.tech/tags/machine-learning.md>), [materials-design](<https://devfeed.tech/tags/materials-design.md>), [optimization](<https://devfeed.tech/tags/optimization.md>), [radix](<https://devfeed.tech/tags/radix.md>), [sparsity](<https://devfeed.tech/tags/sparsity.md>), [time-series](<https://devfeed.tech/tags/time-series.md>)

### AI overview

Sparse Integer Fourier Transform (SIFT) reduces FFT arithmetic by decomposing the Radix-n DFT matrix into four sparse integer matrices and approximating the diagonal matrix of complex twiddle factors through optimization. For radix-32, the authors report a 75% reduction in estimated arithmetic cost, with RMSE below 2.3765x10⁻⁸ and MAPE below 0.01089%. A numerical frequency-extraction test reports MAPE below 2.016% and RMSE below 2.536x10⁻⁶. In wind turbine fault diagnosis, SIFT extracts compact spectral features that retain fault information while reducing classifier input dimensionality compared with the full time series. These findings are reported in the supplied abstract.

### Source excerpt

arXiv:2611.21976v1 Announce Type: cross Abstract: The Discrete Fourier transform, mainly implemented by Fast Fourier Transform (FFT), has been essential for numerous domains, such as digital signal processing and feature extraction in materials design. However, current FFT methods suffer from extensive non-trivial multiplications and twiddle-factor operations in the big data era. To improve the computational efficiency of FFT, this study proposes Sparse Integer Fourier Transform (SIFT), a method that leverages the fast computing of sparsity and small integers, and applies the Radix-n method to derive an FFT. The 2N-point DFT matrix in Radix-n FFT is decomposed into four sparse integer matrices. A diagonal matrix containing complex twiddle factors is efficiently approximated using an optimization strategy while the remaining matrices retain sparsity and integer entries. For radix-32 FFT, SIFT reduces the estimated arithmetic cost by 75%, with RMSE below $2.3765\times10^{-8}$ and MAPE below 0.01089%. We evaluate frequency extraction in a numerical test and an engineering case. In the numerical test, MAPE and RMSE are below 2.016% and $2.536\times10^{-6}$, respectively. For wind turbine fault diagnosis, SIFT extracts compact spectral features that retain fault information while reducing the classifier's input dimensionality relative to the full time series.

## RF-Based Macronutrient Estimation in Liquid Foods via Composition-Aware Multi-Task Learning

DevFeed: [RF-Based Macronutrient Estimation in Liquid Foods via Composition-Aware Multi-Task Learning](<https://devfeed.tech/articles/rf-based-macronutrient-estimation-in-liquid-foods-via-composition-aware-multi-task-learning-137021.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.19052>)

Author: Ishtiaque Ahmed Showmik, Kannan Athreya

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

Content type: article

Language: en

Sources: [arXiv: Networking](<https://devfeed.tech/sources/arxiv-networking.md>)

Topics: [scikit-learn](<https://devfeed.tech/topics/scikit-learn.md>), [fourier transform](<https://devfeed.tech/topics/fourier-transform.md>)

Tags: [2](<https://devfeed.tech/tags/2.md>), [abstract](<https://devfeed.tech/tags/abstract.md>), [analysis](<https://devfeed.tech/tags/analysis.md>), [arxiv](<https://devfeed.tech/tags/arxiv.md>), [based](<https://devfeed.tech/tags/based.md>), [canonical](<https://devfeed.tech/tags/canonical.md>), [categories](<https://devfeed.tech/tags/categories.md>), [challenge](<https://devfeed.tech/tags/challenge.md>), [channel](<https://devfeed.tech/tags/channel.md>), [classes](<https://devfeed.tech/tags/classes.md>), [composition](<https://devfeed.tech/tags/composition.md>), [concatenation](<https://devfeed.tech/tags/concatenation.md>), [condition](<https://devfeed.tech/tags/condition.md>), [content](<https://devfeed.tech/tags/content.md>), [cs-ni](<https://devfeed.tech/tags/cs-ni.md>), [distribution](<https://devfeed.tech/tags/distribution.md>), [domain](<https://devfeed.tech/tags/domain.md>), [driver](<https://devfeed.tech/tags/driver.md>), [learning](<https://devfeed.tech/tags/learning.md>), [liquid](<https://devfeed.tech/tags/liquid.md>), [rf](<https://devfeed.tech/tags/rf.md>)

### AI overview

The study tests whether synchronous 2.4 GHz OFDM radio-frequency measurements can estimate carbohydrate, fat, and protein in liquid foods. Its Ingredient-Conditioned OFDM Transformer (ICOT) processes two-dimensional channel frequency response matrices, infers canonical ingredient categories, then uses those categories to condition macronutrient regression. Across 28,420 captures from 55 beverage classes, it reports mean absolute errors of 0.101 g/100 ml for in-distribution data and 0.485 g/100 ml for out-of-distribution data, described as comparable in magnitude to deviations in manufacturer-reported labels. The authors present this as an early step toward passive, label-free sensing in a liquid-contacting vessel; generalizing to novel formulations remains the primary open challenge.

### Source excerpt

arXiv:2611.19052v1 Announce Type: new Abstract: Current approaches to macronutrient estimation range from high-overhead laboratory analysis to user-dependent methods such as manual nutritional lookup, reading package labels, and vision-based image analysis. These approaches are often time-consuming, input-dependent, or sensitive to imaging and measurement conditions. In this work, we investigate synchronous 2.4 GHz OFDM radio-frequency (RF) measurements for estimating carbohydrate, fat, and protein content in liquid foods. Using a controlled sensing setup, we study how liquid composition shapes frequency-domain RF responses -- a first step toward a sensor integrated into a liquid-contacting vessel that estimates macronutrient content passively and label-free, and toward composition-aware RF food analysis beyond liquids. We propose the Ingredient-Conditioned OFDM Transformer (ICOT), a structure-aware multi-task Transformer that operates on two-dimensional OFDM channel frequency response matrices. ICOT infers canonical ingredient categories representing dielectric-driver composition and uses them to condition macronutrient regression through feature concatenation. On 28,420 captures spanning 55 beverage classes, ICOT achieves an in-distribution MAE of 0.101 g/100 ml and an out-of-distribution MAE of 0.485 g/100 ml -- comparable in magnitude to deviations present in manufacturer-reported labels -- though generalization to novel formulations remains the primary open challenge.

## Vision-Language Model-Based Phase-OTDR Event Classification

DevFeed: [Vision-Language Model-Based Phase-OTDR Event Classification](<https://devfeed.tech/articles/vision-language-model-based-phase-otdr-event-classification-139100.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.20841>)

Author: Ahmet Bestami Kose, Mehmet Sergen Catal, Muhammet Cagri Yeke, Kivilcim Yuksel, Abdurrahman Gumus

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

Content type: article

Language: en

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

Topics: [vlm](<https://devfeed.tech/topics/vlm.md>), [few-shot](<https://devfeed.tech/topics/few-shot.md>), [few-shot-learning](<https://devfeed.tech/topics/few-shot-learning.md>), [image recognition](<https://devfeed.tech/topics/image-recognition.md>), [fourier transform](<https://devfeed.tech/topics/fourier-transform.md>)

Tags: [abstract](<https://devfeed.tech/tags/abstract.md>), [adapter](<https://devfeed.tech/tags/adapter.md>), [arxiv](<https://devfeed.tech/tags/arxiv.md>), [based](<https://devfeed.tech/tags/based.md>), [but](<https://devfeed.tech/tags/but.md>), [cache](<https://devfeed.tech/tags/cache.md>), [class](<https://devfeed.tech/tags/class.md>), [classes](<https://devfeed.tech/tags/classes.md>), [classification](<https://devfeed.tech/tags/classification.md>), [classifier](<https://devfeed.tech/tags/classifier.md>), [clip](<https://devfeed.tech/tags/clip.md>), [code](<https://devfeed.tech/tags/code.md>), [color](<https://devfeed.tech/tags/color.md>), [configuration](<https://devfeed.tech/tags/configuration.md>), [cs-cv](<https://devfeed.tech/tags/cs-cv.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [event](<https://devfeed.tech/tags/event.md>), [language](<https://devfeed.tech/tags/language.md>), [language-model](<https://devfeed.tech/tags/language-model.md>), [model](<https://devfeed.tech/tags/model.md>), [vision](<https://devfeed.tech/tags/vision.md>), [vision-language-model](<https://devfeed.tech/tags/vision-language-model.md>)

### AI overview

The study finds that few-shot adaptation of CLIP visual features can classify Phase-OTDR events with limited labels. It converts raw samples into RGB images of STFT dB-magnitude spectrogram mosaics, then adapts Tip-Adapter-F cache keys while leaving both encoders unchanged. In a leakage-controlled evaluation with 20 support-sampling trials and a fixed held-out test set, ViT-L-14 with OpenAI weights and 128 shots per class achieved 96.81 ± 0.42% accuracy and 96.79 ± 0.42% macro-F1. With 16 shots per class (96 support images across six classes), accuracy was 90.17 ± 0.81%. Tip-Adapter-F outperformed zero-shot CLIP and training-free Tip-Adapter; reported analyses found performance stable across Vision Transformer backbones and prompt wording at the selected setting. The source says code and the image dataset are publicly available.

### Source excerpt

arXiv:2611.20841v1 Announce Type: cross Abstract: Phase-sensitive optical time-domain reflectometry (Phase-OTDR) enables distributed vibration sensing, but reliable event classification has traditionally relied on fully supervised models that require extensive annotated data. This study examines the adaptability of a Vision-Language Model (VLM) for Phase-OTDR event recognition through few-shot learning. Raw samples are converted into Short-Time Fourier Transform (STFT) dB-magnitude spectrogram mosaics, color-mapped and resized to RGB images, and processed by a Contrastive Language-Image Pre-training (CLIP) image encoder; few-shot adaptation optimizes only the Tip-Adapter-F cache keys, leaving both encoders unchanged. Under a leakage-controlled protocol (train-side support/validation pool of 512 images per class, fixed held-out test of 120 per class, no test-set access during selection, 20 distinct support-sampling trials), the selected configuration (the ViT-L-14 Vision Transformer encoder with OpenAI weights, 128 shots per class) reaches 96.81 +/- 0.42% accuracy and a macro-averaged F1 score of 96.79 +/- 0.42% on the held-out test set; a label-efficient 16-shot setting, that is, 96 labeled support images at adaptation time across the six classes, already attains 90.17 +/- 0.81%. Tip-Adapter-F clearly outperforms both zero-shot CLIP and the standard training-free Tip-Adapter, and prompt and backbone analyses show that performance is stable across Vision Transformer encoders and insensitive to prompt wording at the selected operating point. Overall, CLIP visual features carry a transferable structure that a lightweight support cache turns into a label-efficient Phase-OTDR event classifier. The code and the corresponding image dataset are made publicly available on GitHub to support further research: https://github.com/miralab-ai/vlm-phase-otdr.

## FH-Sampling: Frequency-aware Hypergraph-driven Oversampling for Imbalanced Time-Series Classification

DevFeed: [FH-Sampling: Frequency-aware Hypergraph-driven Oversampling for Imbalanced Time-Series Classification](<https://devfeed.tech/articles/fh-sampling-frequency-aware-hypergraph-driven-oversampling-for-imbalanced-time-series-classification-139119.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.20991>)

Author: Yunrong Hao, Ruohao Dai, Zhen Liu, Guan Wang, Wenbin Pei, Qiang Zhang

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

Content type: article

Language: en

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

Topics: [Artificial Intelligence](<https://devfeed.tech/topics/ai.md>), [Temporal data](<https://devfeed.tech/topics/temporal-data.md>), [Time Series](<https://devfeed.tech/topics/time-series.md>), [fourier transform](<https://devfeed.tech/topics/fourier-transform.md>)

Tags: [aircraft](<https://devfeed.tech/tags/aircraft.md>), [applications](<https://devfeed.tech/tags/applications.md>), [classification](<https://devfeed.tech/tags/classification.md>), [components](<https://devfeed.tech/tags/components.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [datasets](<https://devfeed.tech/tags/datasets.md>), [diagnosis](<https://devfeed.tech/tags/diagnosis.md>), [distribution](<https://devfeed.tech/tags/distribution.md>), [experiments](<https://devfeed.tech/tags/experiments.md>), [imbalanced-data](<https://devfeed.tech/tags/imbalanced-data.md>), [sampling](<https://devfeed.tech/tags/sampling.md>), [series](<https://devfeed.tech/tags/series.md>), [techniques](<https://devfeed.tech/tags/techniques.md>), [time](<https://devfeed.tech/tags/time.md>), [time-series](<https://devfeed.tech/tags/time-series.md>), [time-series-data](<https://devfeed.tech/tags/time-series-data.md>)

### AI overview

FH-Sampling addresses imbalanced time-series classification by combining hypergraph-guided sample relationships with frequency-aware synthesis. It builds a hypergraph in the original time-series space to capture higher-order relationships and adaptively assigns sampling ratios to reduce the influence of noisy samples. A discrete wavelet transform separates frequency components; guided by the hypergraph, the method combines high-frequency components from related samples while retaining the target sample's low-frequency component. The abstract reports advantages on 24 imbalanced datasets and says tests on simulated aircraft trajectories produced trajectories that better matched the real distribution. It does not provide metrics or details of the comparisons.

### Source excerpt

arXiv:2611.20991v1 Announce Type: new Abstract: Imbalanced time-series data appears commonly in diverse applications, such as medical diagnosis. When trained on imbalanced data, classifiers tend to exhibit a bias toward the majority class, making it difficult to identify minority-class instances accurately, despite their critical importance. Sampling techniques have proven effective in rebalancing imbalanced data. However, existing interpolation-based oversampling methods typically rely on pairwise interpolation, limiting their ability to capture high-order inter-sample relationships, and their inherent smoothing effects tend to suppress high-frequency components, compromising spectral fidelity. Therefore, we propose a Frequency-aware Hypergraph-driven Oversampling method (FH-Sampling) for imbalanced time-series classification. In FH-Sampling, a hypergraph is first constructed in the original time-series space to capture complex high-order inter-sample relationships, and sampling ratios are adaptively assigned to mitigate the impact of noisy samples. The discrete wavelet transform decomposes each sample into frequency components. Guided by the sample hypergraph, FH-Sampling combines high-frequency components from related samples while retaining the target sample's low-frequency component to help preserve important frequency characteristics. Experiments on 24 imbalanced datasets demonstrate the advantages of FH-Sampling for imbalanced time-series classification. We further apply FH-Sampling to aircraft trajectory generation. Experiments on simulated aircraft trajectories show that FH-Sampling produces trajectories that better match the real distribution, highlighting its potential for practical applications.

## Toward Interference-Agnostic RF Source Separation

DevFeed: [Toward Interference-Agnostic RF Source Separation](<https://devfeed.tech/articles/toward-interference-agnostic-rf-source-separation-135153.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.17950>)

Author: Maxence Albert, Rapha\\"el Duroselle, Thibaud Ehret, Olivier Boeffard, Jean-Fran\\c{c}ois Bonastre

Published: 2026-11-16T05:00:00Z

Content type: article

Language: en

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

Topics: [Decoding](<https://devfeed.tech/topics/decoding.md>), [fourier transform](<https://devfeed.tech/topics/fourier-transform.md>), [Resilience](<https://devfeed.tech/topics/resilience.md>)

Tags: [architectures](<https://devfeed.tech/tags/architectures.md>), [band](<https://devfeed.tech/tags/band.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [eess-sp](<https://devfeed.tech/tags/eess-sp.md>), [rf](<https://devfeed.tech/tags/rf.md>), [separator](<https://devfeed.tech/tags/separator.md>), [signal](<https://devfeed.tech/tags/signal.md>), [source](<https://devfeed.tech/tags/source.md>)

### AI overview

Specialized RF signal separators depend on knowing the signal-interference pair, and the study reports that using one on an unseen interference usually fails, often leaving the signal worse than no processing. The authors train a single separator across all pairs without identifying the interference at test time, using both the ICASSP 2024 RF Challenge WaveNet and an adapted band-split transformer; both show similar results. At the size of one specialized model, the unified separator keeps bit error rates close on nearly every tested pair, performing better on one and worse on another. For an interference absent from all training sets, it matches the best specialized model, without requiring a choice of model at deployment.

### Source excerpt

arXiv:2611.17950v1 Announce Type: new Abstract: Separating a radio-frequency signal of interest from a co-channel interference is mostly addressed with specialized separators, each trained for one pair of signal and interference types and therefore relying on a label that a heavily jammed receiver cannot identify. We show that this reliance is not benign: applied to an interference it was not trained on, such a separator usually fails to recover the signal, and in most of the cases we measure it leaves the signal worse than no processing at all. We then train a single separator over every pair, giving it no indication at test time of which one it faces, on two architectures with no component in common: the convolutional WaveNet of the ICASSP 2024 RF Challenge, and a band-split transformer that we adapt from music source separation. Both behave the same way. At the size of one specialized model, the unified separator stays close to the whole models it replaces, keeping the bit error rate close on almost every pair, better on one and worse on another. On an interference absent from every training set, it does as well as the best specialized model, without anyone having to choose which one to deploy.

## CENDRe: Concept Extraction with Natural Domain Representations

DevFeed: [CENDRe: Concept Extraction with Natural Domain Representations](<https://devfeed.tech/articles/cendre-concept-extraction-with-natural-domain-representations-131954.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2607.29621>)

Author: Antonia Holzapfel, Andres Felipe Posada Moreno, Sebastian Trimpe

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

Content type: article

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: [Artificial Intelligence](<https://devfeed.tech/topics/ai.md>), [Representation learning](<https://devfeed.tech/topics/representation-learning.md>), [fourier transform](<https://devfeed.tech/topics/fourier-transform.md>), [ai observability](<https://devfeed.tech/topics/ai-observability.md>)

Tags: [aggregation](<https://devfeed.tech/tags/aggregation.md>), [automatically](<https://devfeed.tech/tags/automatically.md>), [benchmarks](<https://devfeed.tech/tags/benchmarks.md>), [classification](<https://devfeed.tech/tags/classification.md>), [clusters](<https://devfeed.tech/tags/clusters.md>), [concepts](<https://devfeed.tech/tags/concepts.md>), [contribution](<https://devfeed.tech/tags/contribution.md>), [convolutional-neural-networks](<https://devfeed.tech/tags/convolutional-neural-networks.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [diagnosis](<https://devfeed.tech/tags/diagnosis.md>), [domains](<https://devfeed.tech/tags/domains.md>), [fourier-transform](<https://devfeed.tech/tags/fourier-transform.md>), [localization](<https://devfeed.tech/tags/localization.md>), [masks](<https://devfeed.tech/tags/masks.md>), [predictions](<https://devfeed.tech/tags/predictions.md>), [presence](<https://devfeed.tech/tags/presence.md>), [prototypes](<https://devfeed.tech/tags/prototypes.md>), [signal](<https://devfeed.tech/tags/signal.md>), [space](<https://devfeed.tech/tags/space.md>), [temporal](<https://devfeed.tech/tags/temporal.md>), [time-series](<https://devfeed.tech/tags/time-series.md>)

### AI overview

CENDRe extracts CNN concepts from time-series data and localizes them across signal domains. It clusters per-timestep latent representations in two stages, using silhouette-guided aggregation to select the number of concepts automatically. For localization, gradients of a prototype-based presence score create masks over regions driving each concept; propagated through differentiable invertible input mappings such as the Fourier transform, they also produce localizations in other domains. Per-concept relevance scores estimate contributions to each class. On synthetic benchmarks, CENDRe reports representation correctness comparable to state-of-the-art concept extraction methods and higher importance correctness. On bearing-fault data, it identifies prediction-driving frequency bands in regions commonly inspected for diagnosis, providing evidence unavailable from time-domain-only methods.

### Source excerpt

arXiv:2607.29621v2 Announce Type: replace-cross Abstract: Convolutional neural networks (CNNs) are widely used for time-series classification, but their deployment in critical contexts requires understanding the patterns across signal domains that drive their predictions (e.g., temporal or spectral). Concept extraction (CE) methods identify such patterns by analyzing representations within the models' latent space. However, existing time-series CE methods have three limitations: they operate only in the time domain and cannot locate features in other domains, they depend on a predefined number of concepts that biases them, and they produce localizations misaligned with the regions the model uses. We address these limitations with CENDRe, a concept extraction method for CNNs. For concept discovery, we propose a procedure that clusters per-timestep latent representations in two stages, with a silhouette-guided aggregation that selects the number of concepts automatically. For localization, we define a presence score that contrasts the latent representations with their prototypes, and use its gradients to produce masks that concentrate on the regions driving the concept. These gradients, propagated through a differentiable invertible mapping of the input such as a Fourier transform, yield localizations for the same concepts in different domains. Finally, each concept receives a relevance score that quantifies its contribution to each class. On synthetic benchmarks, CENDRe achieves representation correctness comparable to state-of-the-art CE methods and higher importance correctness, while producing visualizations in different domains. On real bearing-fault data, CENDRe extracts the frequency bands driving the model's predictions, located in regions commonly inspected for fault diagnosis, producing evidence to assess the model that time-domain CE methods cannot.

## Taming Neural Fields with a Frequency-Lipschitz Curriculum for Sparse-View CT Reconstruction

DevFeed: [Taming Neural Fields with a Frequency-Lipschitz Curriculum for Sparse-View CT Reconstruction](<https://devfeed.tech/articles/taming-neural-fields-with-a-frequency-lipschitz-curriculum-for-sparse-view-ct-reconstruction-133020.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.15652>)

Author: Songtao Hu, Jingrui Wu, Hong Ding, Fanhao Zhu, Yunhao Liu, Suyang Xi

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

Content type: article

Language: en

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

Topics: [Image processing](<https://devfeed.tech/topics/image-processing.md>), [mlp](<https://devfeed.tech/topics/mlp.md>), [fourier transform](<https://devfeed.tech/topics/fourier-transform.md>)

Tags: [adaptive](<https://devfeed.tech/tags/adaptive.md>), [capacity](<https://devfeed.tech/tags/capacity.md>), [cs-cv](<https://devfeed.tech/tags/cs-cv.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [curriculum](<https://devfeed.tech/tags/curriculum.md>), [datasets](<https://devfeed.tech/tags/datasets.md>), [details](<https://devfeed.tech/tags/details.md>), [encoding](<https://devfeed.tech/tags/encoding.md>), [mlp](<https://devfeed.tech/tags/mlp.md>), [neural](<https://devfeed.tech/tags/neural.md>), [noise](<https://devfeed.tech/tags/noise.md>), [performance](<https://devfeed.tech/tags/performance.md>), [projection](<https://devfeed.tech/tags/projection.md>), [quality](<https://devfeed.tech/tags/quality.md>), [radiation](<https://devfeed.tech/tags/radiation.md>), [random](<https://devfeed.tech/tags/random.md>), [sampling](<https://devfeed.tech/tags/sampling.md>), [stable](<https://devfeed.tech/tags/stable.md>), [stage](<https://devfeed.tech/tags/stage.md>), [statistics](<https://devfeed.tech/tags/statistics.md>), [training](<https://devfeed.tech/tags/training.md>), [under](<https://devfeed.tech/tags/under.md>)

### AI overview

FLIC controls an implicit neural representation's capacity during sparse-view CT reconstruction with a coarse-to-fine frequency-Lipschitz curriculum. It combines random Fourier feature encoding with a spectrally normalized MLP under scheduled, non-uniform Lipschitz budgets; the final capacity is calibrated from filtered back-projection gradient statistics, avoiding per-scan manual tuning of that budget. Prior-guided adaptive ray sampling places more integration samples near high-gradient anatomical boundaries. The method uses a single training stage without iterative denoising. The abstract reports state-of-the-art unsupervised reconstruction quality across four public CT datasets in noise-free and low-dose settings; it does not provide quantitative results or comparisons in the supplied excerpt.

### Source excerpt

arXiv:2611.15652v1 Announce Type: cross Abstract: Sparse-view computed tomography (SVCT) reduces radiation dose by acquiring fewer projection views, but the resulting under-determined inverse problem requires careful regularization to prevent reconstruction algorithms from fitting streak artifacts and hallucinated structures. Existing implicit neural representation (INR) methods rely on implicit spectral bias or iterative external denoising, yet they provide limited control over when high-frequency capacity is exposed during optimization. To address this limitation, we propose FLIC (Frequency-Lipschitz INR Curriculum), a capacity-controlled self-prior INR framework that makes the network's representational capacity explicit and schedulable. The core of our method is a joint frequency-Lipschitz curriculum that couples random Fourier feature encoding with spectrally normalized MLP decoding under non-uniform Lipschitz budgets, guiding reconstruction along a coarse-to-fine trajectory from stable global anatomy to fine details. The terminal capacity is automatically calibrated from FBP gradient statistics, enabling scan-adaptive regularization without per-scan manual tuning of the terminal Lipschitz budget. We further introduce a prior-guided adaptive ray sampling strategy that concentrates integration samples near high-gradient anatomical boundaries, improving forward-model fidelity. The entire framework operates in a single training stage without iterative denoising, enabling efficient per-image reconstruction. Extensive experiments on four public CT datasets under noise-free and low-dose settings show that our method achieves state-of-the-art unsupervised reconstruction quality with consistent performance across diverse anatomical structures.

## A Time-Frequency-Condition Multimodal Fusion Foundation Model for Universal Rotating-Machinery Fault Diagnosis under Multi-Element Heterogeneous Scenarios

DevFeed: [A Time-Frequency-Condition Multimodal Fusion Foundation Model for Universal Rotating-Machinery Fault Diagnosis under Multi-Element Heterogeneous Scenarios](<https://devfeed.tech/articles/a-time-frequency-condition-multimodal-fusion-foundation-model-for-universal-rotating-machinery-fault-diagnosis-under-multi-element-heterogeneous-scenarios-132089.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.15051>)

Author: Xinyu Zou, Jian Ma, Liang Ma, Chen Lu, Tao Peng, Yingjie Li

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

Content type: article

Language: en

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

Topics: [AI Foundation Models](<https://devfeed.tech/topics/ai-foundation-models.md>), [Machine learning](<https://devfeed.tech/topics/machine-learning.md>), [Large Language Model](<https://devfeed.tech/topics/llm.md>), [llama](<https://devfeed.tech/topics/llama.md>), [few-shot](<https://devfeed.tech/topics/few-shot.md>), [fourier transform](<https://devfeed.tech/topics/fourier-transform.md>)

Tags: [abstract](<https://devfeed.tech/tags/abstract.md>), [adversarial](<https://devfeed.tech/tags/adversarial.md>), [arxiv](<https://devfeed.tech/tags/arxiv.md>), [attention](<https://devfeed.tech/tags/attention.md>), [can](<https://devfeed.tech/tags/can.md>), [class](<https://devfeed.tech/tags/class.md>), [condition](<https://devfeed.tech/tags/condition.md>), [critical](<https://devfeed.tech/tags/critical.md>), [cross](<https://devfeed.tech/tags/cross.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-cl](<https://devfeed.tech/tags/cs-cl.md>), [data](<https://devfeed.tech/tags/data.md>), [datasets](<https://devfeed.tech/tags/datasets.md>), [diagnosis](<https://devfeed.tech/tags/diagnosis.md>), [element](<https://devfeed.tech/tags/element.md>), [foundation](<https://devfeed.tech/tags/foundation.md>), [foundation-model](<https://devfeed.tech/tags/foundation-model.md>), [fusion](<https://devfeed.tech/tags/fusion.md>), [model](<https://devfeed.tech/tags/model.md>), [multimodal](<https://devfeed.tech/tags/multimodal.md>), [time](<https://devfeed.tech/tags/time.md>), [under](<https://devfeed.tech/tags/under.md>)

### AI overview

TFC-MMFD targets fault diagnosis across different rotating machines and operating conditions by combining statistical indicators, time-frequency images, vibration sequences, and textual machine and condition descriptions in a unified sequence. A LLaMA backbone, cross-modal bidirectional attention, and a Q-Former produce fault-mode queries, which are matched to natural-language fault anchors; condition disentanglement and condition-conditioned masked pre-training are intended to improve robustness. Across four tasks on ten public bearing datasets, the abstract reports 99.40% accuracy and 99.16% macro-F1, over 80% accuracy on most interpolated unseen conditions without adaptation, and an increase from 20.26% to 95.63% on an unseen machine with five labeled samples per class. It also reports a 10.57-point gain over the strongest of eight domain-generalization baselines and favorable comparison with a recent bearing foundation model using less than half its fine-tuning data.

### Source excerpt

arXiv:2611.15051v1 Announce Type: cross Abstract: Intelligent fault diagnosis models are typically developed for a single machine under a specific set of operating conditions, such that each new machine or condition requires retraining from scratch. This one-machine-one-model paradigm scales poorly. Although transfer learning and domain generalization can reduce the labeling burden, they primarily seek domain-invariant commonality, potentially discarding machine-specific characteristics that are critical for diagnosis. This paper proposes TFC-MMFD, a time-frequency-condition multimodal fusion foundation model for rotating-machinery fault diagnosis under multi-element heterogeneous scenarios. Statistical indicators, time-frequency images, vibration sequences, and textual machine and condition descriptions are tokenized into a unified language-like space, allowing inputs of different physical origin to be represented as a single sequence. A LLaMA backbone, together with cross-modal bidirectional attention and a Q-Former, then distills this sequence into fault-mode queries. Diagnosis is performed by matching these queries against natural-language fault-mode anchors rather than fixed class indices, while regression-adversarial condition disentanglement and condition-conditioned masked pre-training improve robustness to changing conditions. Across four progressive tasks on ten public bearing datasets spanning diverse bearing types, speeds, and loads, TFC-MMFD attains 99.40% accuracy and 99.16% macro-F1, maintains over 80% accuracy on most interpolated unseen conditions without adaptation, and improves from 20.26% to 95.63% accuracy on an unseen machine with only five labeled samples per class. It outperforms the strongest of eight domain-generalization baselines by 10.57 percentage points and compares favorably with a recent bearing foundation model while using less than half of its fine-tuning data.

## New Learn Guide: ESP-SDR on Metro ESP32-P4 #Hacks #AdafruitLearningSystem @Adafruit

DevFeed: [New Learn Guide: ESP-SDR on Metro ESP32-P4 #Hacks #AdafruitLearningSystem @Adafruit](<https://devfeed.tech/articles/new-learn-guide-esp-sdr-on-metro-esp32-p4-hacks-adafruitlearningsystem-adafruit-129521.md>)

Original publisher: [Read original article](<https://blog.adafruit.com/2026/11/11/new-learn-guide-esp-sdr-on-metro-esp32-p4-hacks-adafruitlearningsystem-adafruit/>)

Author: Adafruit Learning System

Published: 2026-11-11T18:24:38Z

Content type: tutorial

Language: en

Sources: [Adafruit Industries](<https://devfeed.tech/sources/adafruit-industries-makers-hackers-artists-designers-and-engineers.md>)

Topics: [ESP32-C6](<https://devfeed.tech/topics/esp32-c6.md>), [ESP32-P4](<https://devfeed.tech/topics/esp32-p4.md>), [fourier transform](<https://devfeed.tech/topics/fourier-transform.md>), [Learning](<https://devfeed.tech/topics/learning.md>)

Tags: [adafruit-learning-system](<https://devfeed.tech/tags/adafruit-learning-system.md>), [bluetooth](<https://devfeed.tech/tags/bluetooth.md>), [circuitpython](<https://devfeed.tech/tags/circuitpython.md>), [displays](<https://devfeed.tech/tags/displays.md>), [esp](<https://devfeed.tech/tags/esp.md>), [esp-sdr](<https://devfeed.tech/tags/esp-sdr.md>), [esp32](<https://devfeed.tech/tags/esp32.md>), [esp32-c6](<https://devfeed.tech/tags/esp32-c6.md>), [esp32-p4](<https://devfeed.tech/tags/esp32-p4.md>), [espressif](<https://devfeed.tech/tags/espressif.md>), [guide](<https://devfeed.tech/tags/guide.md>), [hacks](<https://devfeed.tech/tags/hacks.md>), [metro](<https://devfeed.tech/tags/metro.md>), [microwave](<https://devfeed.tech/tags/microwave.md>), [radio](<https://devfeed.tech/tags/radio.md>), [sdr](<https://devfeed.tech/tags/sdr.md>), [software-defined-radio](<https://devfeed.tech/tags/software-defined-radio.md>), [waterfall-display](<https://devfeed.tech/tags/waterfall-display.md>), [wifi](<https://devfeed.tech/tags/wifi.md>)

### AI overview

ESP-SDR uses undocumented functionality on the Metro ESP32-P4's ESP32-C6 co-processor to read raw I/Q radio data. The P4 runs CircuitPython and communicates with the C6 over UART; the C6 performs an FFT and sends signal statistics back for a CircuitPython waterfall visualization. Guide pages cover flashing the C6 and loading the P4 project.

### Source excerpt

The ESP-SDR project published a way to interact with undocumented functionality in many ESP32 chips that allows reading raw I/Q data. I/Q stands for In-Phase and Quadrature. It's a digitized arithmetic representation of radio signals captured over time. This page has a more detailed explanation of I/Q data with helpful graphics if you'd like to [...]

## Universal Approximation Theorem for Metaplectic Neural Operators

DevFeed: [Universal Approximation Theorem for Metaplectic Neural Operators](<https://devfeed.tech/articles/universal-approximation-theorem-for-metaplectic-neural-operators-128782.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.13991>)

Author: Ahmed Abdeljawad, Malek Yacine Rezoug

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

Content type: article

Language: en

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

Topics: [Neural Network](<https://devfeed.tech/topics/neural-network.md>), [Machine learning](<https://devfeed.tech/topics/machine-learning.md>), [fourier transform](<https://devfeed.tech/topics/fourier-transform.md>)

Tags: [activation](<https://devfeed.tech/tags/activation.md>), [approximation](<https://devfeed.tech/tags/approximation.md>), [branch](<https://devfeed.tech/tags/branch.md>), [continuous](<https://devfeed.tech/tags/continuous.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [domain](<https://devfeed.tech/tags/domain.md>), [features](<https://devfeed.tech/tags/features.md>), [fixed](<https://devfeed.tech/tags/fixed.md>), [hidden](<https://devfeed.tech/tags/hidden.md>), [layer](<https://devfeed.tech/tags/layer.md>), [machine-learning](<https://devfeed.tech/tags/machine-learning.md>), [math-fa](<https://devfeed.tech/tags/math-fa.md>), [matrix](<https://devfeed.tech/tags/matrix.md>), [neural](<https://devfeed.tech/tags/neural.md>), [operator](<https://devfeed.tech/tags/operator.md>), [operators](<https://devfeed.tech/tags/operators.md>), [space](<https://devfeed.tech/tags/space.md>)

### AI overview

The paper proves compact-uniform universal approximation for metaplectic neural operators on nonempty bounded Lipschitz domains. With a smooth, globally Lipschitz, nonpolynomial activation, one hidden operator layer and one nonlocal branch suffice to approximate continuous operators between C^s and C^r spaces or H^s and H^r spaces, for arbitrary nonnegative integer orders and finite channel dimensions. In the Sobolev case, compact input sets must also be uniformly bounded in L∞. The branch uses a whole-space metaplectic multiplier with zero extension and restriction for a fixed free symplectic matrix; one scalar symbol atom and finitely many positional features are fixed by the transform and domain, independent of the target, input set and accuracy. The proof combines averaging neural operator theory with a chirp-conjugated averaging identity and derivative estimates. The stated results include Fourier and fixed nondegenerate fractional Fourier multiplier constructions.

### Source excerpt

arXiv:2611.13991v1 Announce Type: new Abstract: We prove universal approximation for metaplectic neural operators on nonempty bounded Lipschitz domains. The nonlocal branches are whole-space metaplectic multipliers with zero extension and restriction, associated with any fixed free symplectic matrix. Assuming a smooth, globally Lipschitz, nonpolynomial activation, we obtain compact-uniform approximation of continuous operators from $C^s$ to $C^r$ and from $H^s$ to $H^r$ in the respective output norms, for arbitrary nonnegative integers $s,r$ and finite input and output channel dimensions. In the Sobolev setting, compact input sets must additionally be uniformly bounded in $L^\infty$. One hidden operator layer with one nonlocal branch suffices. A single scalar symbol atom and finitely many positional features are prescribed by the transform and domain independently of the target operator, compact set and accuracy. The proof combines averaging neural operator theory with an exact chirp-conjugated averaging identity and derivative estimates. The result includes Fourier and fixed nondegenerate fractional Fourier multiplier constructions.

## Frequency Domain Reservoir Computing

DevFeed: [Frequency Domain Reservoir Computing](<https://devfeed.tech/articles/frequency-domain-reservoir-computing-128550.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2606.24969>)

Author: Klaus Schertler, Xiomara Runge, Andrea Ceni, David Kappel, Claudio Gallicchio

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

Content type: article

Language: en

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

Topics: [fourier transform](<https://devfeed.tech/topics/fourier-transform.md>), [Computing](<https://devfeed.tech/topics/computing.md>), [Sustainable Computing](<https://devfeed.tech/topics/sustainable-computing.md>)

Tags: [backpropagation](<https://devfeed.tech/tags/backpropagation.md>), [benchmarks](<https://devfeed.tech/tags/benchmarks.md>), [classification](<https://devfeed.tech/tags/classification.md>), [complexity](<https://devfeed.tech/tags/complexity.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [embedding](<https://devfeed.tech/tags/embedding.md>), [energy-consumption](<https://devfeed.tech/tags/energy-consumption.md>), [fixed](<https://devfeed.tech/tags/fixed.md>), [forecasting](<https://devfeed.tech/tags/forecasting.md>), [recurrent](<https://devfeed.tech/tags/recurrent.md>), [reservoir-computing](<https://devfeed.tech/tags/reservoir-computing.md>), [structured](<https://devfeed.tech/tags/structured.md>), [training](<https://devfeed.tech/tags/training.md>)

### AI overview

FRESCO is an Echo State Network architecture designed to reduce the cost of large reservoirs while retaining competitive predictive performance. It uses structured circulant recurrence and stays in the frequency domain to perform dense nonlinear state updates with stated O(N) complexity, avoiding the O(N²) updates that limit conventional ESNs. Dimensional zero-padding for input embedding, a packed frequency-domain readout, and a frequency-domain nonlinearity support this approach. The abstract reports competitive results on memory benchmarks, sequential classification, and multivariate long-horizon forecasting, and claims lower training and inference costs and energy use; it gives no quantitative comparisons.

### Source excerpt

arXiv:2606.24969v2 Announce Type: replace Abstract: While the quadratic sequence-length bottleneck of transformers has fueled a resurgence in recurrent models, effectively capturing complex dynamics requires architectures that balance efficient training with highly expressive latent states. Echo State Networks (ESNs) offer a compelling approach by utilizing fixed recurrent weights to circumvent backpropagation through time, enabling a closed-form training solution. However, achieving the expressivity needed for complex tasks demands large reservoirs, exposing an $\mathcal{O}(N^2)$ state-update bottleneck that prevents ESNs from matching the scale of contemporary recurrent models. To address this limitation, we introduce Frequency Domain Reservoir Computing (FRESCO), an ESN architecture operating entirely in the frequency domain while avoiding domain-shift overheads to achieve $\mathcal{O}(N)$ complexity for dense, non-linear updates via structured circulant recurrence. By employing a novel dimensional zero-padding input embedding, a packed frequency-domain readout, and a natively applied frequency-domain non-linearity, FRESCO drastically reduces computational costs and energy consumption of training and inference. Furthermore, FRESCO achieves competitive predictive performance on memory benchmarks, sequential classification, and multivariate long-horizon forecasting, offering a scalable and energy-efficient path forward for dense recurrent architectures.

## TSFFormer: A Frequency-Domain Spike Transformer for Green Human Activity Recognition

DevFeed: [TSFFormer: A Frequency-Domain Spike Transformer for Green Human Activity Recognition](<https://devfeed.tech/articles/tsfformer-a-frequency-domain-spike-transformer-for-green-human-activity-recognition-125511.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.12090>)

Author: Jiale Deng, Shuiying Xiang

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

Content type: article

Language: en

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

Topics: [Artificial Intelligence](<https://devfeed.tech/topics/ai.md>), [Transformer](<https://devfeed.tech/topics/transformer.md>), [fourier transform](<https://devfeed.tech/topics/fourier-transform.md>), [Neural Network](<https://devfeed.tech/topics/neural-network.md>)

Tags: [2](<https://devfeed.tech/tags/2.md>), [3](<https://devfeed.tech/tags/3.md>), [ablation](<https://devfeed.tech/tags/ablation.md>), [abstract](<https://devfeed.tech/tags/abstract.md>), [arithmetic](<https://devfeed.tech/tags/arithmetic.md>), [art](<https://devfeed.tech/tags/art.md>), [arxiv](<https://devfeed.tech/tags/arxiv.md>), [attention](<https://devfeed.tech/tags/attention.md>), [benchmarks](<https://devfeed.tech/tags/benchmarks.md>), [but](<https://devfeed.tech/tags/but.md>), [complexity](<https://devfeed.tech/tags/complexity.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [deadlock](<https://devfeed.tech/tags/deadlock.md>), [deep](<https://devfeed.tech/tags/deep.md>), [deep-neural-networks](<https://devfeed.tech/tags/deep-neural-networks.md>), [domain](<https://devfeed.tech/tags/domain.md>), [effect](<https://devfeed.tech/tags/effect.md>), [energy](<https://devfeed.tech/tags/energy.md>), [green](<https://devfeed.tech/tags/green.md>), [recognition](<https://devfeed.tech/tags/recognition.md>), [transformer](<https://devfeed.tech/tags/transformer.md>)

### AI overview

TSFFormer is a fully spiking Fourier Transformer for wearable IMU human activity recognition. Its MultiHeadFourier attention mixes information in the frequency domain, with the authors reporting a reduction in complexity from O(V²) to O(V log V) and about 50% lower GPU memory use. A dual-compartment Temporal Spiking LIF node uses dynamic thresholds to adapt to changing activity rhythms and mitigate gradient degradation. Across eight public IMU benchmarks, the authors report that TSFFormer outperforms state-of-the-art SNNs and DeepConvLSTM, trains about three times faster, and has a 0.1 pJ energy footprint. Ablations are said to show benefits from combining the two components.

### Source excerpt

arXiv:2611.12090v1 Announce Type: new Abstract: Wearable Inertial Measurement Units (IMUs) enable pervasive Human Activity Recognition (HAR), but traditional Deep Neural Networks (DNNs) suffer from an "arithmetic-latency deadlock" due to dense floating-point MAC operations. While Spiking Neural Networks (SNNs) offer ultra-low power, existing ones face spatial blind spots, rigid neurons, and gradient degradation. To address these, we propose TSFFormer, a purely spiking Fourier Transformer for Green HAR. It introduces two key innovations: (1) a MultiHeadFourier attention that performs frequency-domain global mixing, reducing complexity from \(O(V^2)\) to \(O(V\log V)\), eliminating spatial blind spots, and saving \(\sim\)50% GPU memory; and (2) a dual-compartment Temporal Spiking LIF (TSLIFNode) that utilizes dynamic thresholds to adapt to non-stationary rhythms and mitigate gradient degradation. Extensive experiments on 8 public IMU benchmarks demonstrate that TSFFormer surpasses state-of-the-art SNNs and even the classic DeepConvLSTM, achieving a \(\sim\)3\(\times\) faster training speed with an energy footprint of 0.1 pJ. Ablation studies confirm the synergistic effect of our modules, providing a highly accurate, ultra-low-power event-driven paradigm for wearable sensing.

## Frequency-Structure Complementary Modeling for Hyperspectral Image Fusion

DevFeed: [Frequency-Structure Complementary Modeling for Hyperspectral Image Fusion](<https://devfeed.tech/articles/frequency-structure-complementary-modeling-for-hyperspectral-image-fusion-125937.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.10413>)

Author: Song Qiya, Tian Siyu, Li Shutao, Dian Renwei, Sun Long, Liu Jinyang

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

Content type: article

Language: en

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

Topics: [Image processing](<https://devfeed.tech/topics/image-processing.md>), [fourier transform](<https://devfeed.tech/topics/fourier-transform.md>), [fourier analysis](<https://devfeed.tech/topics/fourier-analysis.md>)

Tags: [adaptive](<https://devfeed.tech/tags/adaptive.md>), [aggregation](<https://devfeed.tech/tags/aggregation.md>), [attention](<https://devfeed.tech/tags/attention.md>), [benchmarks](<https://devfeed.tech/tags/benchmarks.md>), [channel](<https://devfeed.tech/tags/channel.md>), [coherent](<https://devfeed.tech/tags/coherent.md>), [combined](<https://devfeed.tech/tags/combined.md>), [combines](<https://devfeed.tech/tags/combines.md>), [context](<https://devfeed.tech/tags/context.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-cv](<https://devfeed.tech/tags/cs-cv.md>), [decoder](<https://devfeed.tech/tags/decoder.md>), [dependency](<https://devfeed.tech/tags/dependency.md>), [design](<https://devfeed.tech/tags/design.md>), [fusion](<https://devfeed.tech/tags/fusion.md>), [image](<https://devfeed.tech/tags/image.md>), [interaction](<https://devfeed.tech/tags/interaction.md>), [kernels](<https://devfeed.tech/tags/kernels.md>), [learning](<https://devfeed.tech/tags/learning.md>), [local](<https://devfeed.tech/tags/local.md>), [modeling](<https://devfeed.tech/tags/modeling.md>), [structure](<https://devfeed.tech/tags/structure.md>)

### AI overview

FSCM-Net combines frequency-aware attention with structure-adaptive local refinement for hyperspectral image fusion. Its SSFCA module uses Fourier-interacted query-key representations to form spatial-token affinities and input-conditioned frequency filtering to form latent-channel affinities; a learned shared gain calibrates the channel response before it is combined with the spatial response. Its SC-DFF module uses structure-tensor cues from multispectral features and context-modulated hyperspectral details to produce channel- and position-dependent aggregation kernels, adapting refinement to smooth areas, edges, and textures while aiming to preserve spectral fidelity. The modules are progressively integrated in a multiscale encoder-decoder. The abstract reports favorable spatial-detail reconstruction and spectral-preservation results versus representative methods on two simulated benchmarks and one real remote-sensing dataset; it gives no quantitative results.

### Source excerpt

arXiv:2611.10413v1 Announce Type: cross Abstract: Existing methods mainly rely on convolutional aggregation or attention-based spatial-spectral interaction. However, conventional feature-domain attention does not explicitly characterize variation patterns along spatial-token and learned latent-channel coordinates, while spatially shared local operators lack the flexibility to adapt neighborhood aggregation to heterogeneous structures. To address these limitations, we propose a Frequency-Structure Complementary Modeling Network (FSCM-Net), which couples frequency-aware dependency learning with structure-adaptive local refinement. Specifically, a Spatial-Spectral Frequency Calibration Attention (SSFCA) module constructs spatial-token affinities from Fourier-interacted query-key representations and builds latent-channel affinities after input-conditioned frequency filtering along learned channel coordinates. The latent-channel response is then calibrated by a learned shared gain and combined with the spatial-token response. Complementarily, a Structure-Coherent Dynamic Feature Fusion (SC-DFF) module combines regularized structure-tensor cues from MSI features with context-modulated HSI detail representations to generate channel- and position-dependent aggregation kernels. This design makes local HSI refinement responsive to smooth regions, edges, and textures while maintaining spectral fidelity. The two modules are progressively integrated within a multiscale encoder-decoder to coordinate representation-level interaction and structure-adaptive reconstruction. Experiments on two simulated benchmarks and one real remote-sensing dataset demonstrate favorable performance against representative fusion methods in spatial detail reconstruction and spectral preservation.

## The Fast Tree Transform

DevFeed: [The Fast Tree Transform](<https://devfeed.tech/articles/the-fast-tree-transform-125201.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.10466>)

Author: Paul Burchard

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

Content type: article

Language: en

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

Topics: [fourier transform](<https://devfeed.tech/topics/fourier-transform.md>), [graph theory](<https://devfeed.tech/topics/graph-theory.md>)

Tags: [circle](<https://devfeed.tech/tags/circle.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [fast](<https://devfeed.tech/tags/fast.md>), [graph](<https://devfeed.tech/tags/graph.md>), [math-dg](<https://devfeed.tech/tags/math-dg.md>), [math-sp](<https://devfeed.tech/tags/math-sp.md>), [optimization](<https://devfeed.tech/tags/optimization.md>), [textures](<https://devfeed.tech/tags/textures.md>), [useful](<https://devfeed.tech/tags/useful.md>)

### AI overview

The Fast Tree Transform is a locally aggregated transform for arbitrary graph geometries that generalizes the Hadamard frequency transform and the Haar wavelet transform on the circle. Its pure frequency component approximately diagonalizes the graph Laplacian, making it useful as a preconditioner in geometric optimization. The full transform is only slightly redundant and can represent graph signals containing both edges and textures sparsely.

### Source excerpt

arXiv:2611.10466v1 Announce Type: new Abstract: The Fast Tree Transform is a locally-aggregated transform defined for arbitrary graph geometries that generalizes both the Hadamard frequency transform and the Haar wavelet transform on the circle. The pure frequency part of the transform approximately diagonalizes the graph Laplacian; this fact makes it useful as a preconditioner in geometric optimization problems. The full transform provides an only slightly-redundant representation of signals on the graph that allows sparse representation of signals that include both edges and textures.

## FIND: Frequency Invariance Disentanglement for Test-Time Adaptation in LiDAR 3D Detection

DevFeed: [FIND: Frequency Invariance Disentanglement for Test-Time Adaptation in LiDAR 3D Detection](<https://devfeed.tech/articles/find-frequency-invariance-disentanglement-for-test-time-adaptation-in-lidar-3d-detection-124629.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.09170>)

Author: Dongxiao Li, Zixuan Hu, Ling-Yu Duan

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

Content type: article

Language: en

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

Topics: [3D](<https://devfeed.tech/topics/3d.md>), [fourier transform](<https://devfeed.tech/topics/fourier-transform.md>)

Tags: [3d](<https://devfeed.tech/tags/3d.md>), [adaptive](<https://devfeed.tech/tags/adaptive.md>), [alignment](<https://devfeed.tech/tags/alignment.md>), [architecture](<https://devfeed.tech/tags/architecture.md>), [code](<https://devfeed.tech/tags/code.md>), [consistency](<https://devfeed.tech/tags/consistency.md>), [corruption](<https://devfeed.tech/tags/corruption.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>), [detection](<https://devfeed.tech/tags/detection.md>), [domain](<https://devfeed.tech/tags/domain.md>), [experiments](<https://devfeed.tech/tags/experiments.md>), [features](<https://devfeed.tech/tags/features.md>), [framework](<https://devfeed.tech/tags/framework.md>), [invariance](<https://devfeed.tech/tags/invariance.md>), [invariant](<https://devfeed.tech/tags/invariant.md>), [labeling](<https://devfeed.tech/tags/labeling.md>), [lidar](<https://devfeed.tech/tags/lidar.md>), [test](<https://devfeed.tech/tags/test.md>), [time](<https://devfeed.tech/tags/time.md>)

### AI overview

FIND addresses LiDAR 3D detection under domain shifts by separating domain-invariant content from domain-specific interference in the frequency domain. It uses B-Spline fitting to build locally adaptive filters in a dual-stream architecture: an Invariant Stream extracts features for pseudo-labeling, while a Specific Stream enforces consistency under domain perturbations. The framework combines invariance-driven alignment with stability-guided regularization. The abstract reports that FIND outperforms state-of-the-art methods on cross-dataset adaptation and corruption benchmarks, but provides no quantitative results.

### Source excerpt

arXiv:2611.09170v1 Announce Type: cross Abstract: LiDAR-based 3D Object Detection (L3OD) is a fundamental 3D perception task, yet its performance is often compromised by domain shifts arising from diverse sensor configurations and environmental conditions. Existing Test-Time Adaptation (TTA) methods primarily leverage self-supervision in the spatial domain, yet they frequently confound semantic content with domain style due to the sparse, non-Euclidean nature of point clouds. To overcome this limitation, we propose FIND (Frequency INvariance Disentanglement), a novel TTA framework that precisely disentangles domain-invariant content from domain-specific interference within the frequency domain. Our methodology utilizes B-Spline fitting to construct locally adaptive filters integrated into a dual-stream architecture: an Invariant Stream extracts robust features to guide pseudo-labeling, while a Specific Stream enforces consistency against domain perturbations. By combining invariance-driven alignment with stability-guided regularization, our approach dynamically extracts robust domain-invariant features while suppressing domain-specific interference. Extensive experiments on cross-dataset adaptation and robust corruption benchmarks demonstrate that FIND significantly outperforms state-of-the-art methods. The source code is publicly available at https://github.com/llldx03/FIND.

## Scalable BLE Channel Sounding Localization with Ranging-Assisted PDoA

DevFeed: [Scalable BLE Channel Sounding Localization with Ranging-Assisted PDoA](<https://devfeed.tech/articles/scalable-ble-channel-sounding-localization-with-ranging-assisted-pdoa-123733.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.09451>)

Author: Jorg Wieme, Eli De Poorter, Jeroen Hoebeke

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

Content type: article

Language: en

Sources: [arXiv: Networking](<https://devfeed.tech/sources/arxiv-networking.md>)

Topics: [communications](<https://devfeed.tech/topics/communications.md>), [fourier analysis](<https://devfeed.tech/topics/fourier-analysis.md>), [fourier transform](<https://devfeed.tech/topics/fourier-transform.md>)

Tags: [anchor](<https://devfeed.tech/tags/anchor.md>), [architecture](<https://devfeed.tech/tags/architecture.md>), [ble](<https://devfeed.tech/tags/ble.md>), [bluetooth](<https://devfeed.tech/tags/bluetooth.md>), [body](<https://devfeed.tech/tags/body.md>), [channel](<https://devfeed.tech/tags/channel.md>), [cs-ni](<https://devfeed.tech/tags/cs-ni.md>), [delay](<https://devfeed.tech/tags/delay.md>), [deployment](<https://devfeed.tech/tags/deployment.md>), [eess-sp](<https://devfeed.tech/tags/eess-sp.md>), [error](<https://devfeed.tech/tags/error.md>), [evaluation](<https://devfeed.tech/tags/evaluation.md>), [exchange](<https://devfeed.tech/tags/exchange.md>), [framework](<https://devfeed.tech/tags/framework.md>), [independent](<https://devfeed.tech/tags/independent.md>), [information](<https://devfeed.tech/tags/information.md>), [infrastructure](<https://devfeed.tech/tags/infrastructure.md>), [latency](<https://devfeed.tech/tags/latency.md>), [localization](<https://devfeed.tech/tags/localization.md>), [measurement](<https://devfeed.tech/tags/measurement.md>), [multiple](<https://devfeed.tech/tags/multiple.md>), [number](<https://devfeed.tech/tags/number.md>)

### AI overview

RA-PDoA uses one anchor for a standard two-way Bluetooth Channel Sounding exchange while multiple passive anchors listen to the same transmissions, keeping airtime independent of anchor count. A channel-recovery framework lets passive anchors reconstruct responses for super-resolution delay estimation; coarse arrival-time information bounds position estimates and rejects inconsistent ranges. In modeled indoor simulations with body shadowing, the method reduced 90th-percentile localization error from 175.3 cm to 95.5 cm. The evaluation also examined multipath, synchronization, oscillator impairments, anchor placement, and deployment inaccuracies; the supplied abstract does not give detailed results for those factors.

### Source excerpt

arXiv:2611.09451v1 Announce Type: cross Abstract: Bluetooth Channel Sounding (CS) enables accurate phase-based ranging, but multi-anchor localization requires sequential ranging exchanges that limit scalability and increase measurement latency. This paper proposes Ranging-Assisted Phase Difference of Arrival (RA-PDoA), a scalable BLE localization architecture combining one active anchor performing a standard two-way CS exchange with multiple passive receive-only anchors observing the same transmissions. A channel-recovery framework is derived that enables passive anchors to reconstruct a channel response suitable for super-resolution delay estimation, allowing multipath-robust ranging without participating in a two-way exchange. Coarse arrival-time information is further used to bound the position estimate and reject inconsistent range solutions. A comprehensive evaluation characterizes the impact of multipath propagation, body shadowing, synchronization requirements, oscillator impairments, anchor placement, and deployment inaccuracies, thereby quantifying the practical infrastructure requirements for scalable BLE localization. Simulation results in modeled indoor environments show that RA-PDoA reduces the 90th-percentile localization error from 175.3 cm to 95.5 cm under body shadowing while maintaining an airtime requirement independent of the number of anchors. The results demonstrate that a single active ranging link enables scalable and multipath-robust BLE Channel Sounding localization.

## How Far Can a Trainable Frequency Move? Resolution, Grids and Coupling in Trainable-FrequencyQuantum Models

DevFeed: [How Far Can a Trainable Frequency Move? Resolution, Grids and Coupling in Trainable-FrequencyQuantum Models](<https://devfeed.tech/articles/how-far-can-a-trainable-frequency-move-resolution-grids-and-coupling-in-trainable-frequencyquantum-models-119999.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2602.23409>)

Author: Michael Poppel, Markus Baumann, Sebastian W\\"olckert, Claudia Linnhoff-Popien, Jonas Stein

Published: 2026-11-06T05:00:00Z

Content type: article

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: [fourier transform](<https://devfeed.tech/topics/fourier-transform.md>), [Hyperparameter optimization](<https://devfeed.tech/topics/hyperparameter-optimization.md>), [A/B Testing](<https://devfeed.tech/topics/a-b-testing.md>)

Tags: [circuit](<https://devfeed.tech/tags/circuit.md>), [coupling](<https://devfeed.tech/tags/coupling.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-et](<https://devfeed.tech/tags/cs-et.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [develop](<https://devfeed.tech/tags/develop.md>), [encoding](<https://devfeed.tech/tags/encoding.md>), [gradient](<https://devfeed.tech/tags/gradient.md>), [gradient-flow](<https://devfeed.tech/tags/gradient-flow.md>), [grid](<https://devfeed.tech/tags/grid.md>), [machine-learning](<https://devfeed.tech/tags/machine-learning.md>), [models](<https://devfeed.tech/tags/models.md>), [quant-ph](<https://devfeed.tech/tags/quant-ph.md>), [quantum](<https://devfeed.tech/tags/quantum.md>), [series](<https://devfeed.tech/tags/series.md>)

### AI overview

Trainable-frequency models' movement depends strongly on their starting frequencies. The analysis finds that an isolated real frequency with an optimally adapting coefficient can reach its target under gradient flow when initialized within half a resolution step, while a fixed coefficient may encounter a phase-dependent barrier. A ternary prefactor grid places a model frequency within half a step across a band using logarithmically many encoding gates. In experiments, grid initialization improved synthetic-target fit success from 69% to 84% over the best random start. For targets with many frequencies, however, a fixed grid trained ten times longer outperformed training prefactors. On seven real series, trainable frequencies improved held-out R² by 0.010-0.045 on four, but a classical screen did not distinguish these gains from the others. A classical model with trainable frequencies showed the same dependence on initialization; the results concern training behavior, not a quantum-circuit advantage.

### Source excerpt

arXiv:2602.23409v4 Announce Type: replace-cross Abstract: Trainable-frequency quantum models train the prefactors of their encoding gates so that the circuit's frequencies can move onto the target's. From the unit start of prior work, training rarely reaches a target more than one resolution step away. We show that the target-driven part of the prefactor gradient is a weighted sum of slopes of the overlap $\sin u/u$, where $u$ is the model-target frequency distance scaled by the window half-width. This gives each frequency a reach: under gradient flow, an isolated real frequency whose coefficient adapts optimally reaches its target from within half a step, for every target at least one step from zero frequency; with a fixed coefficient, a phase-dependent barrier can stop it. A ternary prefactor grid places a model frequency within half a step of every band frequency, with logarithmically many encoding gates. Single-frequency circuits trained with Adam and jointly trained amplitudes behave as predicted: inside the first side lobe, a frequency starts toward its target exactly when in phase with it in 98% of runs, and only 11 of 768 frequencies starting beyond the lobe reached their target. With several frequencies, the grid start raised successful fits on synthetic targets from 69% to 84% against the best random start. With many target frequencies, least squares on the fixed grid fits near-perfectly, but the circuit does not reach the coefficients; training the prefactors narrows this shortfall, but with 8 and 24 target frequencies, training the fixed grid ten times longer fitted better. On real series, training the frequencies improved held-out $R^2$ at equal budget by 0.010--0.045 on four of seven series, but a classical screen did not separate these from the others. A classical model with independently trainable frequencies shows the same dependence on the start; our results describe how to train such models, not an advantage of the circuit.

## TACT-KV: Tri-Axis Cosine Transform for Compressing Volumetric KV Caches in Medical VLMs

DevFeed: [TACT-KV: Tri-Axis Cosine Transform for Compressing Volumetric KV Caches in Medical VLMs](<https://devfeed.tech/articles/tact-kv-tri-axis-cosine-transform-for-compressing-volumetric-kv-caches-in-medical-vlms-120635.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.07593>)

Author: Chongyu Qu, Ritchie Zhao, Yufan He, Dong Yang, Zhengyi Lu, Junchao Zhu, Tianyuan Yao, Juming Xiong, Junlin Guo, Yanfan Zhu, Yuechen Yang, Daguang Xu, Bennett Allan Landman, Yucheng Tang, Yuankai Huo

Published: 2026-11-06T05:00:00Z

Content type: article

Language: en

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

Topics: [Compression](<https://devfeed.tech/topics/compression.md>), [Medical imaging](<https://devfeed.tech/topics/medical-imaging.md>), [fourier transform](<https://devfeed.tech/topics/fourier-transform.md>)

Tags: [3d](<https://devfeed.tech/tags/3d.md>), [adaptive](<https://devfeed.tech/tags/adaptive.md>), [bandwidth](<https://devfeed.tech/tags/bandwidth.md>), [batch](<https://devfeed.tech/tags/batch.md>), [benchmarks](<https://devfeed.tech/tags/benchmarks.md>), [bits](<https://devfeed.tech/tags/bits.md>), [cache](<https://devfeed.tech/tags/cache.md>), [compression](<https://devfeed.tech/tags/compression.md>), [context](<https://devfeed.tech/tags/context.md>), [correlation](<https://devfeed.tech/tags/correlation.md>), [cosine](<https://devfeed.tech/tags/cosine.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>), [decoding](<https://devfeed.tech/tags/decoding.md>), [energy](<https://devfeed.tech/tags/energy.md>), [exploits](<https://devfeed.tech/tags/exploits.md>), [feature](<https://devfeed.tech/tags/feature.md>), [fixed](<https://devfeed.tech/tags/fixed.md>), [form](<https://devfeed.tech/tags/form.md>), [medical](<https://devfeed.tech/tags/medical.md>), [vlms](<https://devfeed.tech/tags/vlms.md>)

### AI overview

TACT-KV compresses volumetric vision-language model KV caches by applying a separable 3D discrete cosine transform across the scan's three spatial axes, then assigning each layer's bit budget across frequency bins with reverse water-filling and quantizing bins using a fixed scalar codebook. The authors report the lowest 1-bit prediction-flip rate in 11 of 12 model-dataset pairs across six medical VLMs and two CT benchmarks; they also report a similar advantage on video tasks with four general-purpose VLMs. At equal HBM, the method increased single-GPU decoding batch size by up to 8x. The provided evidence is an abstract, so implementation details and broader evaluation limitations are not available.

### Source excerpt

arXiv:2611.07593v1 Announce Type: cross Abstract: Vision-language models (VLMs) applied to 3D medical scans face a structural memory wall. A single CT or MRI volume produces tens of thousands of vision tokens, and the resulting key-value (KV) cache scales linearly with both batch and context. Existing KV compression methods operate on the cache's feature or token statistics, or apply 1D and 2D transforms, leaving the three-axis spatial structure in the cache unexploited. We propose TACT-KV (Tri-Axis Cosine Transform), a calibration-free KV compression method that exploits strong spatial correlations in visual KV states along all three anatomical axes of 3D scans. This correlation concentrates the KV cache energy under a frequency-domain transform. TACT-KV applies a separable 3D discrete cosine transform to the KV cache, allocates a per-layer bit budget across frequency bins via reverse water-filling (more bits to high-energy components), and quantizes each bin with a fixed scalar codebook. We derive a close-form relation: the distortion reduction from adaptive bit allocation is dominated by the spectral concentration of the transformed cache. Across six medical VLMs and two 3D CT benchmarks, TACT-KV achieves the lowest 1-bit prediction-flip rate on eleven of twelve model and dataset pairs in our main comparison. The same advantage reproduces on four general-purpose VLMs evaluated on video tasks, indicating that the benefit of exploiting three-axis structure generalizes beyond medical imaging. At equal high-bandwidth memory (HBM), TACT-KV scales the decoding batch size on a single GPU by up to $8\times$.

## AdaMamba: Adaptive Frequency-Gated Mamba for Long-Term Time Series Forecasting

DevFeed: [AdaMamba: Adaptive Frequency-Gated Mamba for Long-Term Time Series Forecasting](<https://devfeed.tech/articles/adamamba-adaptive-frequency-gated-mamba-for-long-term-time-series-forecasting-118208.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2604.23239>)

Author: Xudong Jiang, Mingshan Loo, Wengen Li, Mingrui Zhang, Jihong Guan, Shuigeng Zhou

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

Content type: article

Language: en

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

Topics: [Time Series](<https://devfeed.tech/topics/time-series.md>), [Mamba](<https://devfeed.tech/topics/mamba.md>), [fourier transform](<https://devfeed.tech/topics/fourier-transform.md>)

Tags: [adaptive](<https://devfeed.tech/tags/adaptive.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [forecasting](<https://devfeed.tech/tags/forecasting.md>), [mamba](<https://devfeed.tech/tags/mamba.md>), [time-series-forecasting](<https://devfeed.tech/tags/time-series-forecasting.md>)

### AI overview

AdaMamba integrates input-dependent frequency information directly into long-term time-series forecasting. Its interaction learning module captures cross-variable dependencies and multi-scale context; its adaptive frequency-gated state-space module generates context-dependent frequency bases and uses one time-frequency gate to modulate state evolution. This design addresses limitations the abstract identifies in heuristic frequency methods and Mamba forecasters that treat frequency as auxiliary input. The authors report better results than state-of-the-art baselines across seven public benchmarks and four domain-specific datasets, with competitive computational efficiency.

### Source excerpt

arXiv:2604.23239v2 Announce Type: replace Abstract: Time-series forecasting is foundational to a wide spectrum of web-scale services, spanning personalized recommendations, energy scheduling, and infrastructure capacity planning. Accurate long-term time series forecasting (LTSF) requires models to capture long-range temporal dependencies while adapting to periodic patterns that vary across variables. Although frequency-aware methods offer a global view of temporal dynamics, they typically rely on inflexible heuristic designs, such as predefined period lengths, manually selected or randomly sampled top-k frequency components, and separate spectral transformations. Consequently, they fail to account for time-frequency heterogeneity, where variables that appear synchronized in the time domain can differ substantially in the frequency domain. Meanwhile, Mamba-based forecasters, despite their efficiency in long-range modeling, introduce frequency information only as auxiliary signals, leaving spectral evidence disconnected from state transitions. In this work, we propose AdaMamba, a novel approach that explicitly internalizes input-dependent frequency information into the modeling of temporal dependencies. AdaMamba comprises an interaction learning module and an adaptive frequency-gated state space module: the former captures cross-variable dependencies and multi-scale temporal context, while the latter generates context-dependent frequency bases and unifies temporal forgetting into a single time-frequency gate that directly modulates state evolution. Extensive experiments on seven public LTSF benchmarks and four domain-specific datasets demonstrate that AdaMamba consistently outperforms state-of-the-art baselines while maintaining competitive computational efficiency.

## Dual-VaDE: Efficient and Robust Prototype Learning via Variational Dual-Domain Mixture Priors

DevFeed: [Dual-VaDE: Efficient and Robust Prototype Learning via Variational Dual-Domain Mixture Priors](<https://devfeed.tech/articles/dual-vade-efficient-and-robust-prototype-learning-via-variational-dual-domain-mixture-priors-117014.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.05099>)

Author: Wugeng Zheng, Song Wang, Jian Wang

Published: 2026-11-04T05:00:00Z

Content type: article

Language: en

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

Topics: [Machine learning](<https://devfeed.tech/topics/machine-learning.md>), [VAE](<https://devfeed.tech/topics/vae.md>), [fourier transform](<https://devfeed.tech/topics/fourier-transform.md>), [clustering](<https://devfeed.tech/topics/clustering.md>)

Tags: [brain](<https://devfeed.tech/tags/brain.md>), [cluster](<https://devfeed.tech/tags/cluster.md>), [clustering](<https://devfeed.tech/tags/clustering.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [decoder](<https://devfeed.tech/tags/decoder.md>), [em](<https://devfeed.tech/tags/em.md>), [gaussian](<https://devfeed.tech/tags/gaussian.md>), [generative](<https://devfeed.tech/tags/generative.md>), [global](<https://devfeed.tech/tags/global.md>), [inference](<https://devfeed.tech/tags/inference.md>), [local](<https://devfeed.tech/tags/local.md>), [lower-bound](<https://devfeed.tech/tags/lower-bound.md>), [machine-learning](<https://devfeed.tech/tags/machine-learning.md>), [medical](<https://devfeed.tech/tags/medical.md>), [network](<https://devfeed.tech/tags/network.md>), [prototype](<https://devfeed.tech/tags/prototype.md>), [random](<https://devfeed.tech/tags/random.md>), [robust](<https://devfeed.tech/tags/robust.md>), [spatial](<https://devfeed.tech/tags/spatial.md>), [textures](<https://devfeed.tech/tags/textures.md>), [training](<https://devfeed.tech/tags/training.md>)

### AI overview

Dual-VaDE learns interpretable anatomical prototypes from unlabeled medical images by coupling spatial anatomy with phase-preserving Fourier spectral representations in a dual-domain Gaussian mixture prior. It incorporates this prior into the decoder through a closed-form variational lower bound and trains end-to-end with alternating EM, avoiding separate heuristic clustering. On OASIS-3 brain MRI, the abstract reports sharper prototypes, cluster separation improvements of up to 94% over spatial-only baselines, lower variation across random seeds, and 0.28 ms inference overhead per image.

### Source excerpt

arXiv:2611.05099v1 Announce Type: new Abstract: In this paper, we propose Dual-VaDE, an unsupervised generative framework for learning interpretable anatomical prototypes from unlabeled medical images. In contrast to previous variational methods that rely on a single spatial latent code, which often entangles global geometry with local textures and fails to capture their cross-domain dependencies, Dual-VaDE introduces a dual-domain Gaussian mixture prior that explicitly couples spatial anatomy with phase-preserving Fourier spectral representations. We derive an exact, closed-form variational lower bound that fuses this joint prior into the decoder and optimizes the entire network end-to-end via an alternating EM formulation, avoiding heuristic two-stage clustering while ensuring superior training stability and robustness. Evaluated on OASIS-3 brain MRI, Dual-VaDE delivers substantially sharper and more plausible prototypes, achieves superior cluster separation with improvements of up to 94% over spatial-only baselines, exhibits markedly reduced variance across random seeds, and incurs negligible inference overhead (0.28 ms per image).

## Does Source-Free Time-Series Adaptation Really Require Updating the Encoder?

DevFeed: [Does Source-Free Time-Series Adaptation Really Require Updating the Encoder?](<https://devfeed.tech/articles/does-source-free-time-series-adaptation-really-require-updating-the-encoder-116724.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2606.23120>)

Author: Shichang Meng, Linquan Wu, Xuan Ai, Yi Wu, Mingyang Liu, Linqi Song

Published: 2026-11-04T05:00:00Z

Content type: article

Language: en

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

Topics: [Machine learning](<https://devfeed.tech/topics/machine-learning.md>), [Time Series](<https://devfeed.tech/topics/time-series.md>), [fourier transform](<https://devfeed.tech/topics/fourier-transform.md>), [fourier analysis](<https://devfeed.tech/topics/fourier-analysis.md>)

Tags: [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [domain](<https://devfeed.tech/tags/domain.md>), [experiments](<https://devfeed.tech/tags/experiments.md>), [performance](<https://devfeed.tech/tags/performance.md>), [prototype](<https://devfeed.tech/tags/prototype.md>), [time-series](<https://devfeed.tech/tags/time-series.md>), [transformation](<https://devfeed.tech/tags/transformation.md>)

### AI overview

SF-FPA adapts an unlabeled target time series by learning spectral input corrections while keeping the source encoder and classifier frozen. Its frequency adaptation layer adjusts signal magnitudes and phases, using circular phase encoding to avoid angular-boundary discontinuities. Confidence-filtered prototype alignment and information maximization guide the corrections through the frozen predictor toward class-discriminative representations. The abstract reports state-of-the-art average Macro-F1 across three real-world benchmarks; ablations and controlled comparisons with encoder adaptation and a parameter-matched time-domain adapter support the approach. The supplied text does not give benchmark names, scores, or implementation details.

### Source excerpt

arXiv:2606.23120v2 Announce Type: replace Abstract: Source-free domain adaptation (SFDA) adapts a pretrained model to an unlabeled target domain without accessing source data, enabling the reuse of time-series predictors across sensing systems. Existing methods commonly update the encoder using unlabeled target objectives. Yet changes in users, equipment, and operating conditions also alter the spectral characteristics of input signals, raising a fundamental question: does adaptation require changing the encoder, or can we instead adapt the signal it receives? We propose Source Free Frequency-Prototype Alignment (SF-FPA), which learns a spectral input transformation while freezing the source encoder and classifier parameters. First, a frequency adaptation layer applies amplitude and phase corrections to target signals. Separate branches adjust spectral magnitudes and phase relationships, with circular phase encoding avoiding discontinuities at the angular boundary. Second, confidence-filtered prototype alignment guides target representations toward source class prototypes. Combined with information maximization, this objective propagates gradients through the source predictor to learn spectral corrections that support class discrimination. Extensive experiments on three real-world time series benchmarks demonstrate state-of-the-art average Macro-F1 performance. Component ablations and controlled comparisons with encoder adaptation and a parameter-matched time-domain adapter further support spectral input adaptation as an effective approach to reusing source-trained predictors. The implementation is available at https://github.com/ShichangMeng/SF_FPA.

[Next page](<https://devfeed.tech/topics/fourier-transform.md?cursor=WyIyMDI2LTExLTA0VDA1OjAwOjAwKzAwOjAwIiwgIjJhNGU5Yzg4LWM4OGYtNDEwMS05ZjgxLTBjZTdlMjI3NTdlZCJd>)