Multi-modal dialog for browsing large visual catalogs using exploration-exploitation paradigm in a joint embedding space.
Indrani BhattacharyaArkabandhu ChowdhuryVikas C. RaykarPublished in: CoRR (2019)
Keyphrases
- multi modal
- video search
- embedding space
- high dimensional
- exploration exploitation
- low dimensional
- single modality
- dimensionality reduction
- active learning
- visual information
- euclidean space
- manifold learning
- visual features
- data sets
- nonlinear dimensionality reduction
- relevance feedback
- reinforcement learning
- video retrieval
- visual data
- bandit problems
- face recognition