Motion-Guided Masking for Spatiotemporal Representation Learning.
David FanJue WangShuai LiaoYi ZhuVimal BhatHector J. Santos-VillalobosRohith MVXinyu LiPublished in: CoRR (2023)
Keyphrases
- learning process
- space time
- image sequences
- reinforcement learning
- moving objects
- spatio temporal
- prior knowledge
- learning algorithm
- knowledge acquisition
- background knowledge
- spatial and temporal
- supervised learning
- visual representation
- dynamic bayesian networks
- motion patterns
- motion analysis
- learning problems
- active learning
- computer vision