Computer vision
Local proxy that gives vision to text-only models (DeepSeek V4 Flash and any other). A vision model describes your screenshots — text, hex colors, layout — and forwards them as text to your main model. Works with Trae, Cursor, OpenCode, Codex, Claude Code and any OpenAI/Anthropic-compatible client. Dockerized.
Open-source autonomous Tello drone interceptor using YOLO detection, Kalman tracking, visual servoing, target reacquisition, and safety-focused RC control.
MCP server that gives text-only LLM coding agents vision — analyze images via any multimodal model (mimo, Claude, Gemini, OpenAI-compatible). Works with Claude Code, Codex, Kimi, opencode, PI.
Unofficial phone-controlled spray-paint reveal and 3D web experiment
CLI toolkit for AI agents — converts images into structured JSON evidence (OCR, layout, semantics), bridging the vision gap for text-only LLMs.
n8n workflow: Telegram → GPT-4o Vision → Flux → WordPress. Publishes news articles with copyright-safe AI-generated featured images
Give your self-hosted DeepSeek eyes 👀 — vision for DeepSeek-V4-Flash on 2× DGX Spark
Codex skill for local MediaPipe hand tracking and AM-style double-sided vector video effects
Explore computer vision one notebook at a time. GANs · Autoencoders · Style Transfer.
DeepSight: vision-session proxy for text-only LLMs — sketch + targeted tool calls (look/zoom/ocr) with content-addressed caching. OpenAI-compatible.
Give non-vision opencode models the ability to see images by routing them to a vision-capable model.
Synesthesia Encoder (通感编码器) — give any text-only LLM (DeepSeek, etc.) the ability to see images via structured spatial text encoding. A Pi agent extension.
An advanced Computer Vision & AI application in Python that allows you to play Subway Surfers by controlling your character directly with real-time body movements and hand gestures captured via your webcam.
Give a text-only coding agent a pair of eyes. Portable Agent Skill that reads images through the Codex CLI.
Agent skill: 本地 macOS OCR/classify/saliency via Apple Vision — 不上云、零费用、隐私安全
视频 or 图片深度参考图创作应用
High-fidelity, bare-metal industrial simulation environment written in Nim to generate Data Matrix (ECC200) datasets with real-world physical defects for YOLO-OBB training