Weakly Supervised Training of Universal Visual Concepts for Multi-domain Semantic Segmentation.
Petra BevandicMarin OrsicJosip SaricIvan GrubisicSinisa SegvicPublished in: Int. J. Comput. Vis. (2024)
Keyphrases
- weakly supervised
- semantic segmentation
- object categories
- object detectors
- superpixels
- training examples
- topic models
- object class
- relation extraction
- object detection
- object classes
- semi supervised
- conditional random fields
- training samples
- named entities
- object recognition
- supervised learning
- data sets
- generative model
- training set
- support vector
- small number
- object segmentation
- natural images
- domain specific
- image features