#video understanding
4 resultsAdversarial Seeded Sequence Growing for Weakly-Supervised Temporal Action Localization
Chengwei Zhang, Yunlu Xu, Zhanzhan Cheng +4
The paper proposes a weakly‑supervised framework for temporal action localization that alternates between a Seeded Sequence Growing network, which expands reliable seed regions, an…
Learning Visual Actions Using Multiple Verb-Only Labels
Michael Wray, Dima Damen
The paper proposes using multiple verb-only labels for video action recognition and cross‑modal retrieval, learning these labels via regression to capture semantic ambiguities and…
Multi-Granularity Fusion Network for Proposal and Activity Localization: Submission to ActivityNet Challenge 2019 Task 1 and Task 2
Haisheng Su, Xu Zhao, Shuming Liu
The paper introduces a Multi-Granularity Fusion Network that merges proposals from various top‑down and bottom‑up frameworks to improve temporal action proposal generation and loca…
Submission to ActivityNet Challenge 2019: Task B Spatio-temporal Action Localization
Chunfei Ma, Joonhyang Choi, Byeongwon Lee +1
The paper describes an end‑to‑end RGB‑only system for spatio‑temporal action localization in videos, built on SlowFast networks and enhanced with correlation‑preserving data augmen…