Plugging Self-Supervised Monocular Depth into Unsupervised Domain Adaptation for Semantic Segmentation.
Adriano CardaceLuca De LuigiPierluigi Zama RamirezSamuele SaltiLuigi di StefanoPublished in: WACV (2022)
Keyphrases
- semantic segmentation
- street scenes
- conditional random fields
- superpixels
- label transfer
- image sequences
- weakly supervised
- scene classification
- object categories
- pascal voc
- object class
- object classes
- depth information
- depth estimation
- object detection
- depth map
- object recognition
- image understanding
- information extraction
- three dimensional
- input image
- multiple objects
- object segmentation
- image set
- bag of words