Native macOS Computer Use MCP server with a live agent cursor, Apple Vision OCR, fail-closed targeting, and Accessibility actions.
Intelligent video analysis powered by OpenAI Vision API. Extract frames, analyze content, search videos semantically, and estimate token costs with FFmpeg and GPT-4/GPT-5 models.
š MCP server for image analysis using Cloudflare Workers AI (Llama 3.2 Vision) - Analyze images with structured JSON output
Give AI agents eyes and hands on macOS ā screenshot, OCR, click, type through the Model Context Protocol
Browser-based OCR SDK powered by PaddleOCR, ONNX Runtime Web and OpenCV.js
Face and pupil detection library
Manage Roboflow computer vision projects and image datasets. Discover projects and image batches, upload and annotate images, search and tag datasets, generate and export dataset versions, start and stop model training, list trained models, run hosted obj
A Javascript library for computer vision in the browser
Headless node.js test runner for Repeato mobile UI tests
šŖ Darbot Windows MCP - Professional Windows desktop automation server for AI agents via Model Context Protocol (MCP). Control Windows like a human with 15 powerful tools. Includes Python dependencies, UV, and VS Code configuration.
Create a computer vision endpoint in seconds
Image tracing and vectorization plugin for genart.dev ā edge detection, contour extraction, Bezier spline fitting, stipple/hatching, threshold presets, multi-pass refinement
MnemoPay agent toolkit ā the capability layer (runtime, agents, browser, computer-control, vision, voice, scrape, knowledge, tools, coding-agent, research-agent) that pairs with @mnemopay/sdk's portable trust layer. One install for the full agent stack.
MCP server for AI-powered image and video analysis - supports OpenAI, Claude, and multimodal vision APIs with local file and URL processing
Powerful React image and video annotation tool for machine learning, computer vision, and AI training data creation. Features rotatable bounding boxes, polygons, keypoints, segmentation with TypeScript support and Material-UI components.
High-performance React Native SDK for image pre-processing for ML/AI pipelines. Native implementations for pixel extraction, tensor conversion, quantization, augmentation, and model-specific preprocessing (YOLO, MobileNet, etc.)
Object detection in the browser. 100% MIT Licensed. Supports YOLOX, YOLO9, RF-DETR, and YOLOv8/v11/v26 via ONNX Runtime (WebGPU + WASM).
Computer Vision Library - Reactive Extensions
Use your YOLO onnx object detection model in Typescript Bun environment easily.
VisionCraft MCP Server - AI Knowledge Base for Claude AI, Cursor and Windsurf
A lightweight, zero-dependency React document scanner with real-time edge detection, stability filtering, and perspective correction (bilinear warping) for high-resolution A4 captures.
Intuned SDK computer-use agent: drive a desktop through xdotool with a vision-capable model
MCP server for AI vision analysis via OpenRouter
High-performance, jitter-free image tracking for the mobile web ā a WASM+SIMD AR engine
Browser-native spatial 3D digital sculpting and animation studio engine with real-time gesture tracking, GPU-accelerated vertex deformation, and adaptive One Euro Filter signal smoothing.
A new algorithm for retrieving topological skeleton as a set of polylines from binary images: WASM version
On-device React component for real-time auto-capture of ID cards and passports. Detection and capture run entirely in the browser via onnxruntime-web.
OpenCV compiled to WebAssembly for browsers
Advanced MCP server for intelligent computer use automation and AI agent operator control. Supports mouse and keyboard input, window-relative vision and coordinates, Retina display, and screenshot-based debugging. Designed for AI-driven workflows, app and
3D hand tracking in the browser with no special hardware: one webcam, several cameras (including phones over WebRTC), or WebXR headsets, with guided calibration.
Privacy-first AI skills for React applications. Zero backend, zero costs, 100% privacy. Run summarization, translation, speech recognition, and computer vision entirely in the browser.
Use head tracking and face gestures to move a "mouse" pointer from up to 3m (10ft) away š
Cloud Annotations models for TensorFlow.js
Node-RED wrapper arround Microsoft Azure Cognitive Face API
JavaScript SDK for CompreFace - free and open-source face recognition system from Exadel
Engine-agnostic 2D vision core: line-of-sight, field-of-view cones, and visibility-polygon (shadow-casting) computation against segment/rect/polygon obstacles. The gameplay-level visibility layer Cocos (and most engines) don't ship ā stealth detection, AI
Local UI-grounding specialist for hybrid AI agents. Screenshot + text target ā strict JSON bbox. Qwen3-VL-2B LoRA, MLX 4-bit + GGUF + Ollama builds. Daemon, HTTP server, batch, confidence, eval. Drop-in for Claude Code, Codex, browser-use, Skyvern. Cuts G
A-Frame wrappers for Zappar's computer vision library
npm CLI shim for the Fovux MCP Python package.
Control an Android phone from OpenCode: live phone screen in the TUI sidebar, agent tools for taps/typing/swipes/UI dumps, app-specific skills, and a built-in phone-vision subagent. Works over USB, wireless, or Tailscale.
Core computer vision logic and mathematical utilities for Active Liveness Detection.