video-object-tracking 开源项目推荐（16 个）

gaomingqi /Track-Anything

Track-Anything is a flexible and interactive tool for video object tracking and segmentation, based on Segment Anything, XMem, and E2FGVI.

SimonZeng7108 /efficientsam3

EfficientSAM3通过渐进式知识蒸馏将SAM3压缩为轻量级、边缘友好型模型，实现快速可提示的概念分割和跟踪。

JunweiLiang /Object_Detection_Tracking

用于 CMU 的多摄像头监控视频对象检测和跟踪系统的开箱即用代码和模型。速度优化的 Faster-RCNN 模型。基于 Tensorflow。也支持 EfficientDet。WACVW'20

Zhongdao /UniTrack

[NeurIPS'21] Unified tracking framework with a single appearance model. It supports Single Object Tracking (SOT), Video Object Segmentation (VOS), Multi-Object Tracking (MOT), Multi-Object Tracking and Segmentation (MOTS), Pose Tracking, Video Instance Segmentation (VIS), and class-agnostic MOT (e.g. TAO dataset).

royshil /obs-detect

Object Detection in OBS, real-time, local, GPU optional

nerminsamet /houghnet

[ECCV-20] Official PyTorch implementation of HoughNet, a voting-based object detector.

Zongwei97 /UnTrack

[CVPR 2024] Unified Multi-Sensor Tracker With One Parameter Set

ojh6404 /deep_vision_ros

ROS package for SOTA Computer Vision Models including SAM, Cutie, GroundingDINO, YOLO-World, VLPart, DEVA and MaskDINO.

Avinash793 /object-tracking-software

Implementation of Lucas Kanade Tracking system using six parameter affine model and recursive Gauss-Newton process.

eshoyuan /TrackGPT

TrackGPT: Track What You Need in Videos via Text Prompts

inspirai /Characterize-Anything

Characterize Anything: A Wondrous Chemical Reaction between vision models and AI Characters

nerminsamet /houghnet-vid

[TPAMI-22] Bottom-up, voting based video object detection method

Youhanna-Yousry /Tracking-Objects-in-Videos

This project explores Lucas-Kanade algorithm to track an object in a video sequence

ojh6404 /VisionAnything

Python package for SOTA computer vision tasks.

hieulenho /grounded-sam-2-research-benchmark

Reproducible Grounded-SAM-2 custom-video benchmarks with metrics, reports, and annotated outputs

Fujiry0 /stvg-agent

Training-free code-as-action agent for spatio-temporal video grounding (STVG): Qwen3.5-9B writes and runs Python that calls SAM 3 tracking and VLM queries. Evaluated on AnyGroundBench.