Understanding the properties and limitations of contrastive learning for Out-of-Distribution detection.
Nawid KeshtmandRaúl Santos-RodríguezJonathan LawryPublished in: CoRR (2022)
Keyphrases
- supervised learning
- learning algorithm
- detection algorithm
- learning process
- learning systems
- neural network
- probability distribution
- online learning
- learning problems
- detection method
- external representations
- subject matter
- elementary school
- deeper understanding
- learning tasks
- anomaly detection
- knowledge acquisition
- object detection
- active learning
- prior knowledge
- reinforcement learning