# MediaPipe Published articles for MediaPipe. This is one page of public article previews, not the complete archive. Follow Next page to continue. Summaries are not the original full articles. ## Building a Real-Time 3D Face Mask with MediaPipe, Threlte and Three.js DevFeed: [Building a Real-Time 3D Face Mask with MediaPipe, Threlte and Three.js]() Original publisher: [Read original article]() Author: Marek Jóźwiak Published: 2026-09-06T13:09:52Z Content type: tutorial Language: en Sources: [Codrops]() Topics: [Three.js webcam]() Tags: [3d](), [3d-face-mask](), [canonical-face-model](), [creative-coding](), [face-landmarks](), [face-mesh](), [face-tracking](), [facial-landmarks](), [generative-graphics](), [gpu](), [inference](), [interactive-3d](), [javascript-3d](), [mediapipe](), [mediapipe-face-landmarker](), [mesh](), [real-time](), [real-time-3d](), [real-time-face-tracking](), [svelte](), [three-js](), [three-js-face-mask](), [three-js-webcam](), [threlte](), [threlte-three-js](), [tutorial](), [tutorials](), [video](), [vision](), [wasm](), [webcam](), [webcam-effects](), [webcam-face-tracking](), [webgl](), [webgl-face-tracking]() ### AI overview Tutorial on building a real-time textured 3D face mask by mapping MediaPipe face landmarks onto a Three.js mesh with Threlte. ### Source excerpt Learn how MediaPipe's face landmarks, Google's canonical face model, and Three.js come together to create a real-time textured 3D face mask. ## Building an AI Tennis Coach with MediaPipe and Claude DevFeed: [Building an AI Tennis Coach with MediaPipe and Claude]() Original publisher: [Read original article]() Published: 2026-02-25T00:00:00Z Content type: tutorial Language: en Sources: [Gaurav Sarma's Blog]() Topics: [Streamlit](), [MediaPipe Face Landmarker](), [Claude](), [Artificial Intelligence](), [Code](), [math]() Tags: [ai](), [app](), [building](), [code](), [floating-point](), [laptop](), [math](), [mediapipe](), [pipeline](), [speed]() ### AI overview This tutorial explains how to build a Streamlit AI tennis-coaching app using MediaPipe pose detection and Claude. The pipeline analyzes uploaded tennis video, calculates joint angles and swing timing, overlays a skeleton on the footage, and generates coaching feedback grounded in the computed measurements. ### Source excerpt . [Building an AI Tennis Coach with MediaPipe and Claude](building-an-ai-tennis-coach-mediapipe-claude-cover... ## Arm & ExecuTorch 0.7: Bringing Generative AI to the masses DevFeed: [Arm & ExecuTorch 0.7: Bringing Generative AI to the masses]() Original publisher: [Read original article]() Author: EricSondhi; Gian Marco Iodice Published: 2025-08-13T14:55:10Z Content type: article Language: en Sources: [Hugging Face - Blog]() Topics: [Arm](), [Generative AI](), [Artificial Intelligence](), [Large Language Model](), [AI Inference](), [cpu](), [Mobile](), [Raspberry Pi](), [Latency]() Tags: [ai](), [ai-accelerator](), [android](), [arm](), [cpu](), [cross-platform](), [developers](), [edge-ai](), [generative](), [generative-ai](), [large-language-models-llms](), [latency](), [llama-cpp](), [mediapipe](), [memory](), [onnx](), [performance](), [raspberry-pi](), [smartphones]() ### AI overview Arm's KleidiAI will be enabled by default in the upcoming ExecuTorch 0.7 beta, providing automatic AI acceleration through ExecuTorch and XNNPack on current and older Arm-based devices. The article explains how this can improve model startup time, latency, and memory efficiency, making Generative AI and large language models more practical on mobile devices and Raspberry Pi 5. It also describes how the Arm SDOT instruction accelerates low-precision matrix multiplication used by LLMs. ### Source excerpt With Arm's recent SME2 announcement, the role of Arm KleidiAI is increasingly clear as Arm's AI accelerator layer powering the next wave of AI. By embedding into widely-used Edge AI frameworks like XNNPack, MediaPipe, MNN, ONNX Runtime, and even llama.cpp, KleidiAI has delivered substantial performance improvements with no code changes required by developers. ## Getting Started with Gemma 3 LLM on Android for On-Device Inference DevFeed: [Getting Started with Gemma 3 LLM on Android for On-Device Inference]() Original publisher: [Read original article]() Author: blundell Published: 2025-03-12T22:11:14Z Content type: tutorial Language: en Sources: [Blundell]() Topics: [gemma](), [On-device AI](), [Large Language Model](), [Android](), [Inference](), [MediaPipe](), [App](), [Fine-tuning]() Tags: [android](), [androiddev](), [fine-tuning](), [gemma](), [inference](), [intermediate](), [llm](), [mediapipe](), [on-device](), [reference](), [tutorial]() ### AI overview A tutorial for integrating the Gemma 3 1B open-weight language model into an Android app for on-device inference. It covers downloading the model after installation, using MediaPipe for inference, and the trade-offs involving model size, performance, and device compatibility. ### Source excerpt This post gets you started with the new Gemma 3 model for on-device inference. Giving you the simplest steps to get started with AI on Android. Gemma 3 1B is a new model size in the Gemma family of open weight models. When deploying small language models (SLMs) in production settings, models need to be small [...] The post Get started with Gemma 3 LLM on Android now! first appeared on Blundell. ## AI Image Classification for PEZ Collectors | Vertex AI & MediaPipe on Android DevFeed: [AI Image Classification for PEZ Collectors | Vertex AI & MediaPipe on Android]() Original publisher: [Read original article]() Author: Mike Wolfson Published: 2024-03-26T18:00:00Z Content type: tutorial Language: en Sources: [My Big Appetite - Mike Wolfson]() Topics: [Artificial Intelligence](), [data](), [App](), [Mobile](), [Website](), [MediaPipe](), [Google]() Tags: [ai](), [android](), [app](), [classification](), [data](), [dataset](), [devices](), [google](), [image-classification](), [mediapipe](), [mobile](), [pez](), [tech](), [vertex](), [vertex-ai]() ### AI overview This tutorial describes building an image-classification model to distinguish visually similar PEZ dispensers. It covers creating a labeled photo dataset, training the model with Vertex AI, and using MediaPipe to support integration on smartphones or websites. ### Source excerpt I've always had a soft spot for PEZ dispensers and have been collecting them for over 30 years. These quirky little collectibles come in an incredible variety of shapes and characters. But there's more to PEZ than just fun; some dispensers can be quite valuable depending on their age and variation. To address the challenge of identification, I harnessed the power of AI to create an image classification model that could help identify these subtle differences. Use Case: Identifying PEZ dispensers Identifying the precise type of PEZ dispenser isn't always easy. Take Mickey Mouse dispensers, for example. While even a novice collector can broadly identify it as Mickey, subtle differences in the face or eyes could separate a common version from a prized rarity. A $1 Mickey and a $150 Mickey can look awfully similar. To illustrate the problem, these Mickey Mouse dispensers look similar, but are worth drastically different amounts (from left-to-right). Mickey Die-Cut (1961) - $125 Mickey B (1989) - $15 Mickey C (1997) - $1 It is fairly easy for even an untrained human to tell the difference between these dispensers. There are obvious differences between the shape of the face, and the eyes. Could a Computer Tell the Difference? I decided to see if I could train a custom image classification model to distinguish between PEZ dispensers. That model, in theory, could be embedded into a mobile app or a website, giving PEZ enthusiasts a tool to aid in identification. For this project, I embraced technologies from Google: Vertex AI to manage my image dataset and train the model, and MediaPipe for easy integration onto edge devices like smartphones. Step 1: Picture This - Building the Dataset Like any good AI project, I needed data. My original plan was to scrape images from the web, but there simply weren't enough consistent, high-quality pictures for the level of detail I wanted. The best solution? Become a PEZ photographer! I gathered a dozen dispensers from my collection and met ## MobileDiffusion: Rapid text-to-image generation on-device DevFeed: [MobileDiffusion: Rapid text-to-image generation on-device]() Original publisher: [Read original article]() Author: Google AI (noreply@blogger.com) Published: 2024-01-31T21:59:00Z Content type: article Language: en Sources: [Google Research]() Topics: [Machine learning](), [Mobile](), [text-to-image](), [Inference](), [model-deployment](), [Optimization](), [Android](), [iOS](), [MediaPipe]() Tags: [android](), [computer-vision](), [deployment](), [diffusion](), [generation](), [image](), [inference](), [ios](), [machine-learning](), [mediapipe](), [mobile](), [on-device](), [on-device-learning](), [optimization](), [text-to-image]() ### AI overview Google researchers introduce MobileDiffusion, a latent diffusion model designed for on-device text-to-image generation. Using DiffusionGAN for one-step sampling, it was tested on premium iOS and Android devices and generated 512x512 images in about half a second with a 520-million-parameter model. ### Source excerpt Posted by Yang Zhao, Senior Software Engineer, and Tingbo Hou, Senior Staff Software Engineer, Core ML Text-to-image diffusion models have shown exceptional capabilities in generating high-quality images from text prompts. However, leading models feature billions of parameters and are consequently expensive to run, requiring powerful desktops or servers (e.g., Stable Diffusion, DALL-E, and Imagen). While recent advancements in inference solutions on Android via MediaPipe and iOS via Core ML have been made in the past year, rapid (sub-second) text-to-image generation on mobile devices has remained out of reach. To that end, in "MobileDiffusion: Subsecond Text-to-Image Generation on Mobile Devices", we introduce a novel approach with the potential for rapid text-to-image generation on-device. MobileDiffusion is an efficient latent diffusion model specifically designed for mobile devices. We also adopt DiffusionGAN to achieve one-step sampling during inference, which fine-tunes a pre-trained diffusion model while leveraging a GAN to model the denoising step. We have tested MobileDiffusion on iOS and Android premium devices, and it can run in half a second to generate a 512x512 high-quality image. Its comparably small model size of just 520M parameters makes it uniquely suited for mobile deployment. Rapid text-to-image generation on-device. Background The relative inefficiency of text-to-image diffusion models arises from two primary challenges. First, the inherent design of diffusion models requires iterative denoising to generate images, necessitating multiple evaluations of the model. Second, the complexity of the network architecture in text-to-image diffusion models involves a substantial number of parameters, regularly reaching into the billions and resulting in computationally expensive evaluations. As a result, despite the potential benefits of deploying generative models on mobile devices, such as enhancing user experience and addressing emerging privacy concer