Skip to content

Official implementation for ACM MM'24 paper: Open-Vocabulary Video Scene Graph Generation via Union-aware Semantic Alignment

Notifications You must be signed in to change notification settings

ZiyueWu59/UASAN

Repository files navigation

UASAN

Official implementation for ACM MM'24 paper: Open-Vocabulary Video Scene Graph Generation via Union-aware Semantic Alignment

1.Download data and pre-trained models from followling link(Link,Password: MM24)

2. Put all data in /OpenVidSGG folder(as the dataset folder), and put the pre-trained models in /UASAN/ folder.

3. Run tools/eval_traj_cls_both.py for the results in Table 1, and run tools/eval_relation_cls.py for the results in Table 2 and 3.

Acknowledgement

This repo contains modified codes from OpenVoc-VidVRD. We sincerely thank the owners of the great repo!

About

Official implementation for ACM MM'24 paper: Open-Vocabulary Video Scene Graph Generation via Union-aware Semantic Alignment

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages