Casting Geometric Constraints in Semantic Segmentation as Semi-Supervised Learning.
Sinisa StekovicFriedrich FraundorferVincent LepetitPublished in: CoRR (2019)
Keyphrases
- geometric constraints
- semi supervised learning
- semantic segmentation
- semi supervised
- superpixels
- weakly supervised
- conditional random fields
- unlabeled data
- multiple images
- labeled data
- supervised learning
- supervised methods
- epipolar geometry
- unsupervised learning
- machine learning
- object class
- image understanding
- object categories
- active learning
- scene classification
- training data
- object classes
- object segmentation
- single image
- visual features
- object recognition
- pascal voc
- feature selection
- computer vision