Towards Fine-grained Visual Representations by Combining Contrastive Learning with Image Reconstruction and Attention-weighted Pooling.
Jonas DippelSteffen VoglerJohannes HöhnePublished in: CoRR (2021)
Keyphrases
- fine grained
- image reconstruction
- coarse grained
- reinforcement learning
- learning process
- access control
- learning algorithm
- learning experience
- learning analytics
- emission computed tomography
- image reconstruction algorithms
- sparsely sampled
- visual representations
- visual representation
- super resolution
- high resolution
- feature extraction
- image processing