Technology
MediaPipe
Google's open-source, cross-platform framework for building and deploying real-time machine learning pipelines across mobile, web, and edge devices.
MediaPipe is a powerful, open-source framework from Google for applied machine learning (ML). It enables developers to construct complex, multimodal pipelines using a graph-based dataflow model: data moves through connected 'Calculators' for efficient, real-time processing of video, audio, and text . The core strength is its cross-platform capability: build once, and deploy seamlessly across Android, iOS, web (JavaScript), desktop, and IoT devices (like Coral) . MediaPipe offers ready-to-use 'Solutions' (Tasks) for critical computer vision applications, including Face Detection, Hand Tracking, and Pose Estimation (detecting 33 key body points) . This streamlined approach accelerates ML integration, minimizing the need for extensive model tuning or infrastructure setup .
What builders pair with MediaPipe
Projects using both technologies. Select a pairing to see a project.
12 more pairings
Pairing: OpenCV
Super Coool Super Awesome Tennis Serve Analyzer
Pairing: PyTorch
Exercise Posture Tracking Using Computer Vision
Pairing: AI
Building Emotional Intelligence for AI
Pairing: Android
Using AI to channel Olympic excitement into skill development of homegrown athletes
Pairing: Autodesk MotionBuilder
Virtual avatar generation models as world navigators
Pairing: Autoencoder
Exercise Posture Tracking Using Computer Vision
Recent Talks & Demos
Showing 1-7 of 7