Computer vision
Computer vision projects read images and video rather than generate them, answering questions about what a picture shows, spotting objects, or working out where a photo was taken. LLaVA and Ask-Anything show the pattern: a vision model paired with a language model so the agent can describe scenes and hold a conversation about them.
430 agents · most starred first
Top 100 of 430
[NeurIPS'23 Oral] Visual Instruction Tuning (LLaVA) built towards GPT-4V level capabilities and beyond.
An agent skill that finds where a photo was taken — OpenStreetMap geometry, elevation skylines, satellite imagery and street view — and shows its work. Works with Claude Code, Codex, Cursor, Gemini CLI, OpenCode and GitHub Copilot.
超好用的截图工具
[CVPR2024 Highlight][VideoChatGPT] ChatGPT with video understanding! And many more supported LMs such as miniGPT4, StableLM, and MOSS.
InternGPT (iGPT) is an open source demo platform where you can easily showcase your AI models. Now it supports DragGAN, ChatGPT, ImageBind, multimodal chat like GPT-4, SAM, interactive image editing, etc. Try it at igpt.opengvlab.com (支持DragGAN、ChatGPT、ImageBind、SAM的在线Demo系统)
Open-Source AI Camera Skills Platform, AI NVR & CCTV Surveillance. Local VLM video analysis with Qwen, DeepSeek, SmolVLM, LLaVA, YOLO26. LLM-powered agentic security camera agent — watches, understands, remembers & guards your home via Telegram, Discord or Slack. Pluggable AI skills. OpenAI, Google, Anthropic or local AI. Runs on Mac Mini & AI PC.
A cross-platform video structuring (video analysis) framework based on CV models & mLLM.
Cambrian-1 is a family of multimodal LLMs with a vision-centric design.
NVIDIA AI Blueprint for video search and summarization (VSS) is a GPU-accelerated reference architecture for building video analytics agents with real-time verified alerts, visual Q&A, and automated reporting. The VSS Blueprint uses vision language models (VLMs) such as NVIDIA Cosmos, LLMs such as NVIDIA Nemotron, RAG, and NVIDIA NIMs.
Must-have resource for anyone who wants to experiment with and build on the OpenAI vision API 🔥
[ACL 2024 🔥] Video-ChatGPT is a video conversation model capable of generating meaningful conversation about videos. It combines the capabilities of LLMs with a pretrained visual encoder adapted for spatiotemporal video representation. We also introduce a rigorous 'Quantitative Evaluation Benchmarking' for video-based conversational models.
Eyes for text-only DeepSeek Harness agents: built-in free vision chain (no key) + pixel-level vision tools (Q&A, grounding, crop, pixel diff, colors, OCR, SVG trace, cutout, screenshots). One-command install, no Python, image turns work like ordinary tool-calling turns.
为纯文本模型"看图“设计更好的视觉工具箱和技能,支持多图理解,图片问答,前端UI还原、GUI 自动化等,并可选无缝接入多个主流agent,直接识别粘贴图片| A vision toolkit and skill designed for text-only llms — image Q&A, long-screenshot OCR, frontend UI restoration, and GUI automation, with optional seamless integration for Codex, Claude Code, Pi, Oh My Pi, and OpenCode
[CVPR 2024 🔥] Grounding Large Multimodal Model (GLaMM), the first-of-its-kind model capable of generating natural language responses that are seamlessly integrated with object segmentation masks.
Official repo for MM-REACT
[NeurIPS 2025] 4KAgent: Agentic Any Image to 4K Super-Resolution. An intelligent computer vision agent that can magically restore any image to perfect-4K!
Use the reMarkable2 as an interface to vision-LLMs (ChatGPT, Claude, Gemini). Ghost in the machine!
Image to text, fast.
🔍 Search local images with natural language on Android, powered by OpenAI's CLIP model. / 在 Android 上用自然语言搜索本地图片 (基于 OpenAI 的 CLIP 模型)
一个面向长视频内容理解的 Video Agent 平台。
Contrastive Language-Image Forensic Search allows free text searching through videos using OpenAI's machine learning model CLIP
[CVPR 2025 Highlight] Official code for "Olympus: A Universal Task Router for Computer Vision Tasks"
[CVPR 2024 Highlight] OPERA: Alleviating Hallucination in Multi-Modal Large Language Models via Over-Trust Penalty and Retrospection-Allocation
transformers pytorch qwen text-generation custom_code zh
🧠 VideoMind: A Chain-of-LoRA Agent for Temporal-Grounded Video Reasoning (ICLR 2026)
The visual communication layer between humans and AI agents. Capture, annotate, render diagrams, and organize with AI — powered by Electron and Ollama. macOS & Linux.
[CVPR2024] ViP-LLaVA: Making Large Multimodal Models Understand Arbitrary Visual Prompts
[ICRA 2024] Chat with NeRF enables users to interact with a NeRF model by typing in natural language.
This is the official code of VideoAgent: A Memory-augmented Multimodal Agent for Video Understanding (ECCV 2024)
Official Repository of paper VideoGPT+: Integrating Image and Video Encoders for Enhanced Video Understanding
Lightweight GPT-4 Vision processing over the Webcam
CLIP⚡NCNN⚡基于自然语言的图片搜索(Image Search)⚡以字搜图⚡x86⚡Android
[CVPR 2026] LongVT: Incentivizing "Thinking with Long Videos" via Native Tool Calling
Code/Data for the paper: "LLaVAR: Enhanced Visual Instruction Tuning for Text-Rich Image Understanding"
(AAAI 2024) BLIVA: A Simple Multimodal LLM for Better Handling of Text-rich Visual Questions
General video interaction platform based on LLMs, including Video ChatGPT
AI assistant that can query visual datasets, search the FiftyOne docs, and answer general computer vision questions
AgentSTAR: Agentic Shape Tracking and Reconstruction from Monocular Videos
OmniFusion — a multimodal model to communicate using text and images
CDAF (Cached Descriptive Asset Files) - open sidecar format for video so AI agents stop re-analyzing the same footage. Spec, CLI, agent skill, reproducible benchmark.
transformers safetensors text-generation chat cogvlm2 conversational
openai-captcha-detection 是一个使用 OpenAI 进行验证码识别的工具。目前验证码识别准确率100%,通过调用 OpenAI 的 API,这个项目可以实现对复杂验证码图片的文本识别,帮助开发者在验证码处理场景中进行自动化操作。openai-captcha-detection is a tool that uses OpenAI for captcha recognition. Currently, it achieves a 100% accuracy rate in captcha recognition.
Agent4FaceForgery: Multi-Agent LLM Framework for Realistic Face Forgery
Sort a folder of images according to their similarity with provided text in your browser (uses a browser-ported version of OpenAI's CLIP model and the web's new File System Access API)
transformers safetensors text-generation custom_code en arxiv:2311.03079
Context Encoding for Semantic Segmentation MegaDepth: Learning Single-View Depth Prediction from Internet Photos LiteFlowNet: A Lightweight Convolutional Neural Network for Optical Flow Estimation PWC-Net: CNNs for Optical Flow Using Pyramid, Warping, and Cost Volume On the Robustness of Semantic Segmentation Models to Adversarial Attacks SPLATNet: Sparse Lattice Networks for Point Cloud Processing Left-Right Comparative Recurrent Model for Stereo Matching Enhancing the Spatial Resolution of Stereo Images using a Parallax Prior Unsupervised CCA Discovering Point Lights with Intensity Distance Fields CBMV: A Coalesced Bidirectional Matching Volume for Disparity Estimation Learning a Discriminative Feature Network for Semantic Segmentation Revisiting Dilated Convolution: A Simple Approach for Weakly- and Semi- Supervised Semantic Segmentation Unsupervised Deep Generative Adversarial Hashing Network Monocular Relative Depth Perception with Web Stereo Data Supervision Single Image Reflection Separation with Perceptual Losses Zoom and Learn: Generalizing Deep Stereo Matching to Novel Domains EPINET: A Fully-Convolutional Neural Network for Light Field Depth Estimation by Using Epipolar Geometry FoldingNet: Interpretable Unsupervised Learning on 3D Point Clouds Decorrelated Batch Normalization Unsupervised Learning of Depth and Egomotion from Monocular Video Using 3D Geometric Constraints PU-Net: Point Cloud Upsampling Network Real-Time Monocular Depth Estimation using Synthetic Data with Domain Adaptation via Image Style Transfer Tell Me Where To Look: Guided Attention Inference Network Residual Dense Network for Image Super-Resolution Reflection Removal for Large-Scale 3D Point Clouds PlaneNet: Piece-wise Planar Reconstruction from a Single RGB Image Fully Convolutional Adaptation Networks for Semantic Segmentation CRRN: Multi-Scale Guided Concurrent Reflection Removal Network DenseASPP: Densely Connected Networks for Semantic Segmentation
An ever-growing playground of notebooks showcasing CLIP's impressive zero-shot capabilities
Open-source video processing core and CLI. Search and annotate local videos with your own model endpoints.
[MTI-LLM@NeurIPS 2025] Official implementation of "PyVision: Agentic Vision with Dynamic Tooling."
[EMNLP2025]Official implementation: Agent-style vision question answer in Autonomous Driving!
Local-first YOLO workbench for annotation, training, evaluation, and AI assistance.
Implementation of PALI3 from the paper PALI-3 VISION LANGUAGE MODELS: SMALLER, FASTER, STRONGER"
Agent skill: outline, mask and count objects in photos with Meta's SAM 3 on your own NVIDIA GPU, with review, exports and a viewer. Works with Claude Code, Codex, Cursor, Pi and other Agent Skills clients.
Video surveilance footage analyst powered by GPT
[MM 2024] [Need only a 3090] MiniGPT-3D: Efficiently Aligning 3D Point Clouds with Large Language Models using 2D Priors
Our idea is to combine the power of computer vision model and LLMs. We use YOLO, CLIP and DINOv2 to extract high-level features from images. We pass the prompt, along with the extracted features, to LLM, allowing for advanced image dataset queries.
🔥[CVPR2025] EventGPT: Event Stream Understanding with Multimodal Large Language Models
[ICCV 2025] The official code of the paper "Deciphering Cross-Modal Alignment in Large Vision-Language Models with Modality Integration Rate".
GPT-4 Vision Chrome Extension
[CVPR2024] This is the official implement of MP5
[CVPR 2025 🔥]A Large Multimodal Model for Pixel-Level Visual Grounding in Videos
Agent skill for deepfake detection & media safety — detect AI-generated audio, images, and video with Resemble AI
Multimodal chatbot with computer vision capabilities integrated, our 1st-gen LMM
transformers safetensors qwen text-generation custom_code zh
VGGT 3D Vision Agent optimized for Apple Silicon with Metal Performance Shaders
[ECCV 2026] Promsa: Progressive Multimodal Search Agents for Knowledge-Based Visual Question Answering
一个面向长视频内容理解的 Video Agent 平台。
transformers safetensors gemma4 image-text-to-text bf16 multimodal
Give your AI agent a face. Four full-screen visualizers. A living circuit board with a cinematic fly-through, a particle-orb starburst, Matrix rain that looks back, and a constellation brain. All driven live by your voice agent's signal bus. Zero dependencies: a browser and one tiny Python server. Pairs with backtalk out of the box.
OmniAgent (ICML 2026): the first native omni-modal agent for active video perception — a 7B agent that beats Qwen2.5-VL-72B with 73% fewer frames on LVBench.
[ACM Multimedia 2025] "Multi-Agent System for Comprehensive Soccer Understanding"
Arion is the first agentic AI desktop app for geospatial analysis, developed by GeoRetina Inc.
Multi-modal Critical Thinking Agent Framework for Complex Visual Reasoning
A simple package to use CLIP on apple silicon using the MLX libraries from Apple
Early Alpha Release: Chat with Your Image - Leveraging GPT-4 Vision and Function Calls for AI-Powered Image Analysis and Description
FreeVA: Offline MLLM as Training-Free Video Assistant
transformers safetensors text-generation custom_code en arxiv:2312.08914
Vailabel Studio - a local-first desktop studio that labels your data with an offline AI copilot. Private by default. Windows, macOS, Linux.
Image and video processing shaders for ReShade.
Agentic AI platform transforming passive CCTV into proactive railway safety intelligence. Real-time behavioural detection (suicide risk, pickpocketing) via YOLOv8 + BiLSTM + multi-agent reasoning — zero new hardware, privacy-first, no facial recognition.
Extract information, summarize, ask questions, and search videos using OpenAI's Vision API 🚀🎦
OpenAI CLIP converted to Tensorflow 2/Keras
SmartLabel AI 2026: Auto-Annotate Any Object via LLM-Powered Prompt Parsing
Multimodal Instruction Tuning for Llama 3
PyTorch implementation of "SkeletonAgent: An Agentic Interaction Framework for Skeleton-based Action Recognition"
Private local AI Photographer Agent on AMD Radeon and ROCm
[NeurIPS'25] ReAgent-V: A Reward-Driven Multi-Agent Framework for Video Understanding
A demo for running quantized CLIP model (ViT-B/32) on Android.
The official repo of VideoAgentTrek
Plugins for Agent DVR
Analyze any video and produce an .avt (Agentic Video Transcript) file — structured visual descriptions, timestamped transcripts, scene tags, and frame references for AI agent consumption.
Agent-ready Skills for Roboflow and Computer Vision workflows
[Arxiv] Video-RSI: Recursive Self-Improvement of Video Understanding Agents via Harness Evolution
[NeurIPS DB 2025] IR3D-Bench: Evaluating Vision-Language Model Scene Understanding as Agentic Inverse Rendering
Code for "Agentic Very Long Video Understanding" (EGAgent) [ACL 2026 Main]
[IEEE TMM] Light-Omni: Reflex over Reasoning in Agentic Video Understanding with Long-Term Memory
transformers pytorch llava text-generation region:us
transformers safetensors llama text-generation multimodal vision-language
A utility that leverages GPT-4o to rename image files based on their content
Learning to Detect Objects from Multi-Agent LiDAR Scans without Manual Labels. (CVPR2025)