yihongXU/DAUMOT
Official Implementation for DAUMOT: Domain Adaptation for Unsupervised Multiple Object Tracking, An unsupervised MOT training framework with domain adaptation.
This project helps computer vision engineers develop more robust multiple object tracking (MOT) systems. It takes existing MOT models trained on one dataset and adapts them to perform well on new, unseen video footage without requiring manual re-labeling of the new data. Researchers and engineers working on autonomous vehicles, surveillance, or robotics would use this to improve tracking performance across diverse environments.
No commits in the last 6 months.
Use this if you need to deploy an object tracking system in a new visual environment where you lack labeled training data, but have an existing model trained on different data.
Not ideal if you have ample labeled data for your target domain or if you are not working with multiple object tracking tasks.
Stars
12
Forks
1
Language
—
License
GPL-3.0
Category
Last pushed
Mar 14, 2022
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/computer-vision/yihongXU/DAUMOT"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
roboflow/rf-detr
[ICLR 2026] RF-DETR is a real-time object detection and segmentation model architecture...
stereolabs/zed-sdk
⚡️The spatial perception framework for rapidly building smart robots and spaces
mikel-brostrom/boxmot
BoxMOT: Pluggable SOTA multi-object tracking modules with support for axis-aligned and oriented...
RizwanMunawar/yolov7-object-tracking
YOLOv7 Object Tracking Using PyTorch, OpenCV and Sort Tracking
google-deepmind/tapnet
Tracking Any Point (TAP)