-
[pdf]
[arXiv]
[bibtex]@InProceedings{Nguyen-Nhu_2025_CVPR, author = {Nguyen-Nhu, Tinh-Anh and Tran, Huu-Loc and Le, Nguyen-Khang and Nguyen, Minh-Nhat and Nguyen, Tien-Huy and Nguyen-Huu, Hoang-Long and Phan-Nguyen, Huu-Phong and Pham, Huy-Thach and Nguyen, Quan and Le, Hoang M. and Dinh, Quang-Vinh}, title = {A Lightweight Moment Retrieval System with Global Re-Ranking and Robust Adaptive Bidirectional Temporal Search}, booktitle = {Proceedings of the Computer Vision and Pattern Recognition Conference (CVPR) Workshops}, month = {June}, year = {2025}, pages = {3708-3718} }
A Lightweight Moment Retrieval System with Global Re-Ranking and Robust Adaptive Bidirectional Temporal Search
Abstract
The exponential growth of digital video content has posed critical challenges in moment-level video retrieval, where existing methodologies struggle to efficiently localize specific segments within an expansive video corpus. Current retrieval systems are constrained by computational inefficiencies, temporal context limitations, and the intrinsic complexity of navigating video content. In this paper, we address these limitations through a novel Interactive Video Corpus Moment Retrieval framework that integrates a SuperGlobal Reranking mechanism and Adaptive Bidirectional Temporal Search (ABTS), strategically optimizing query similarity, temporal stability, and computational resources. By preprocessing a large corpus of videos using a keyframe extraction model and deduplication technique through image hashing, our approach provides a scalable solution that significantly reduces storage requirements while maintaining high localization precision across diverse video repositories.
Related Material