XML_RVMR

This repository contains the XML model for the baseline of the Ranked Video Moment Retrieval (RVMR) task. The associated paper is titled "Video Moment Retrieval in Practical Setting: A Dataset of Ranked Moments for Imprecise Queries."

The main repository of the paper is TVR-Ranking, and this model is adapted from TVRetrieval.

Annotations and features can be downloaded from TVR-Ranking. The environment setup is the same as for RelocNet_RVMR, as detailed in the TVR-Ranking repository.

Performance

Model Train Set Top N IoU=0.3 IoU=0.5 IoU=0.7
Val Test Val Test Val Test
NDCG@10
XML 1 0.1016 0.0917 0.0747 0.0660 0.0244 0.0268
XML 20 0.2226 0.2135 0.1623 0.1567 0.0580 0.0627
XML 40 0.2002 0.2044 0.1461 0.1502 0.0541 0.0589
NDCG@20
XML 1 0.1010 0.0923 0.0737 0.0662 0.0258 0.0269
XML 20 0.2331 0.2243 0.1700 0.1650 0.0627 0.0664
XML 40 0.2114 0.2167 0.1530 0.1590 0.0583 0.0635
NDCG@40
XML 1 0.1077 0.1016 0.0775 0.0727 0.0273 0.0294
XML 20 0.2580 0.2512 0.1874 0.1853 0.0705 0.0753
XML 40 0.2408 0.2432 0.1740 0.1791 0.0666 0.0720

Quick Start

Modify the path in run_top20.sh and then execute the script:

sh run_top20.sh

Feel free to contribute or raise issues for any problems encountered.

Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model is not currently available via any of the supported Inference Providers.
The model cannot be deployed to the HF Inference API: The model has no library tag.

Dataset used to train LiangRenjie/XML_RVMR