Looking Beyond Single Images for Contrastive Semantic Segmentation Learning.
Feihu ZhangPhilip H. S. TorrRené RanftlStephan R. RichterPublished in: NeurIPS (2021)
Keyphrases
- semantic segmentation
- street scenes
- image database
- learning algorithm
- input image
- image data
- supervised learning
- weakly supervised
- image features
- image retrieval
- image understanding
- bounding box
- three dimensional
- active learning
- high resolution
- object recognition
- segmentation algorithm
- image regions
- image set
- superpixels
- pixel level
- training data
- image segmentation