co-tracker
Point tracker
A model for tracking any point on a video using transformer-based architecture and optical flow benefits
CoTracker is a model for tracking any point (pixel) on a video.
4k stars
34 watching
262 forks
Language: Jupyter Notebook
last commit: almost 2 years agooptical-flowpoint-trackingtrack-anything
Related projects:
| Repository | Description | Stars |
|---|---|---|
 visionml/pytracking | A comprehensive framework for building and training visual object tracking models using PyTorch. | 3,271 |
 doubiiu/dynamicrafter | This project generates animated videos from open-domain images by leveraging pre-trained video diffusion priors. | 2,668 |
| facebookresearch/segment-anything | This project provides code and tools for running inference with a visual segmentation model that can generate object masks from input prompts. | 48,092 |
| facebookresearch/sam2 | A software framework for video segmentation in images and videos using AI models | 13,054 |
 eduardolundgren/tracking.js | A computer vision library for the web that provides various algorithms and techniques for tracking objects and features in images. | 9,445 |
| facebookresearch/slowfast | Provides state-of-the-art video understanding codebase with efficient training methods and pre-trained models for various tasks | 6,680 |
 clementpinard/sfmlearner-pytorch | Pytorch implementation of unsupervised depth and ego-motion learning from video sequences | 1,022 |
| doubiiu/tooncrafter | Generates cartoon-style videos from two images using pre-trained diffusion models | 5,447 |
| facebookresearch/dinov2 | A PyTorch implementation of a self-supervised learning method for learning robust visual features without supervision. | 9,425 |
 mkocabas/vibe | A video pose and shape estimation method that predicts body parameters for each frame of an input video. | 2,911 |
| facebookresearch/detectron2 | A platform for object detection and segmentation tasks using machine learning algorithms | 30,778 |
 benaclejames/vrcfacetracking | Provides eye and lip tracking functionality for VRChat avatars using OSC protocol and custom parameter mappings | 634 |
 thudm/cogvideo | Generates videos from text and images using large language models | 9,761 |
| facebookresearch/imagebind | An AI framework that combines data from multiple sources into a single embedding space, enabling various applications such as cross-modal retrieval and generation. | 8,424 |
 martinruenz/co-fusion | Enables a robot to maintain scene descriptions at the object level by segmenting and tracking objects in real-time from RGB-D images. | 502 |