DART routes zero-shot video temporal grounding queries by difficulty using DPP entropy, achieving up to 3.5 mIoU gains with 7x fewer frames on Charades-STA and ActivityNet Captions.
arXiv preprint arXiv:1909.00239 (2019)
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
cs.CV 2years
2026 2verdicts
UNVERDICTED 2representative citing papers
MCMT improves weakly-supervised VMR by fusing multiple learnable Gaussian masks from proposals into a positive sample mask and using dual masked query reconstruction tasks for stability.
citing papers explorer
-
DART: Difficulty-Adaptive Routing for Zero-Shot Video Temporal Grounding
DART routes zero-shot video temporal grounding queries by difficulty using DPP entropy, achieving up to 3.5 mIoU gains with 7x fewer frames on Charades-STA and ActivityNet Captions.
-
Multi-proposal Collaboration and Multi-task Training for Weakly-supervised Video Moment Retrieval
MCMT improves weakly-supervised VMR by fusing multiple learnable Gaussian masks from proposals into a positive sample mask and using dual masked query reconstruction tasks for stability.