A two-stage framework improves viewpoint-invariant and temporally consistent action detection in untrimmed videos by using virtual viewpoint augmentation at training time and a selective state-space temporal encoder, outperforming prior methods on PKU-MMD and BABEL.
Harnessing temporal causal- ity for advanced temporal action detection
5 Pith papers cite this work. Polarity classification is still indexing.
5
Pith papers citing it
years
2026 5representative citing papers
LiquidTAD distills liquid neural dynamics into a vectorized parallel temporal operator and hierarchical decay sharing to achieve efficient action detection with substantially reduced model size and computation.
SkillSpotter raises class-specific mAP from 12.40 to 21.82 and balanced accuracy to 60.40% on Ego-Exo4D by adding adaptive temporal suppression, gated pose fusion, and bidirectional cross-view attention to temporal action detectors.
EgoAction uses decoupled verb-noun temporal detectors on VideoMAE features and Dynamic Weighted Fusion of boundaries based on classification confidences for the EPIC-KITCHENS action detection challenge.
citing papers explorer
-
Improving Viewpoint-Invariance and Temporal Consistency for Action Detection
A two-stage framework improves viewpoint-invariant and temporally consistent action detection in untrimmed videos by using virtual viewpoint augmentation at training time and a selective state-space temporal encoder, outperforming prior methods on PKU-MMD and BABEL.
-
LiquidTAD: Efficient Temporal Action Detection via Parallel Liquid-Inspired Temporal Relaxation
LiquidTAD distills liquid neural dynamics into a vectorized parallel temporal operator and hierarchical decay sharing to achieve efficient action detection with substantially reduced model size and computation.
-
SkillSpotter: Pose-Aware Multi-View Skilled Action Detection and Grading in Ego-Exo Videos
SkillSpotter raises class-specific mAP from 12.40 to 21.82 and balanced accuracy to 60.40% on Ego-Exo4D by adding adaptive temporal suppression, gated pose fusion, and bidirectional cross-view attention to temporal action detectors.
-
EgoAction: Egocentric Action Composition with Reliability-Aware Temporal Fusion for the EPIC-KITCHENS Action Detection Challenge at CVPR 2026
EgoAction uses decoupled verb-noun temporal detectors on VideoMAE features and Dynamic Weighted Fusion of boundaries based on classification confidences for the EPIC-KITCHENS action detection challenge.
- Event-Causal RAG: A Retrieval-Augmented Generation Framework for Long Video Reasoning in Complex Scenarios