S2DNet: Learning Image Features for Accurate Sparse-to-Dense Matching.
Hugo GermainGuillaume BourmaudVincent LepetitPublished in: ECCV (3) (2020)
Keyphrases
- learning process
- image features
- learning algorithm
- reinforcement learning
- dictionary learning
- prior knowledge
- learning systems
- sparse data
- learning problems
- computationally efficient
- image representation
- knowledge acquisition
- online learning
- feature vectors
- feature points
- object detection
- active learning
- sparse representation
- mobile learning
- matching algorithm
- stereo matching
- support vector
- neural network