Transformer-Based Semantic Segmentation for Extraction of Building Footprints from Very-High-Resolution Images.
Jia SongA-Xing ZhuYunqiang ZhuPublished in: Sensors (2023)
Keyphrases
- semantic segmentation
- street scenes
- image data
- input image
- image database
- ground truth
- image features
- label transfer
- image retrieval
- image classification
- information extraction
- multiple images
- image set
- three dimensional
- test images
- image understanding
- image matching
- conditional random fields
- image collections
- object categories
- automatic extraction
- pixel level
- d objects
- viewpoint
- computer vision