OVIS: Open-Vocabulary Visual Instance Search via Visual-Semantic Aligned Representation Learning.
Sheng LiuKevin LinLijuan WangJunsong YuanZicheng LiuPublished in: CoRR (2021)
Keyphrases
- visual learning
- visual representation
- learning process
- visual features
- learning algorithm
- perceptual information
- visual perception
- search algorithm
- search space
- visual processing
- semantic search
- visual information
- high level
- semantic representation
- visual search
- low level
- online learning
- video search
- intermediate representations
- bayesian networks
- reinforcement learning
- visual representations
- visual vocabulary
- prior knowledge
- domain independent
- semantic web
- visual concepts
- image representation
- learning systems
- spatial relations
- search strategy
- background knowledge