VDSM: Unsupervised Video Disentanglement with State-Space Modeling and Deep Mixtures of Experts.
Matthew J. VowelsNecati Cihan CamgözRichard BowdenPublished in: CoRR (2021)
Keyphrases
- state space
- video sequences
- video frames
- unsupervised manner
- spatial and temporal
- real time
- search space
- video content
- video streams
- video database
- unsupervised learning
- state variables
- deep architectures
- mixtures of gaussians
- gaussian mixture
- event detection
- dynamical systems
- space time
- video data
- supervised learning
- reinforcement learning
- multimedia
- learning algorithm
- human activities
- human actions
- video surveillance
- video retrieval
- blind source separation
- online video