Shape-biased CNNs are Not Always Superior in Out-of-Distribution Robustness.
Xinkuan QiuMeina KanYongbin ZhouYanchao BiShiguang ShanPublished in: WACV (2024)
Keyphrases
- shape model
- shape representation
- probability distribution
- high quality
- shape recognition
- shape retrieval
- shape analysis
- fixed point
- shape features
- density function
- medial axis
- shape recovery
- graphical models
- computer vision
- cellular neural networks
- arbitrarily shaped
- shape descriptors
- shape prior
- random variables
- dissimilarity measure
- aspect ratio
- shape change