Scene-Aware Deep Networks for Semantic Segmentation of Images.
Zhike YiTao ChangShuai LiRuijun LiuJing ZhangAimin HaoPublished in: IEEE Access (2019)
Keyphrases
- semantic segmentation
- street scenes
- input image
- multiple images
- three dimensional
- image set
- scene classification
- image database
- superpixels
- label transfer
- image data
- image classification
- ground truth
- image collections
- image understanding
- image regions
- image retrieval
- image features
- multiple objects
- pixel level
- segmentation algorithm
- test images
- single image
- target object
- object recognition
- depth estimation
- bounding box
- spatial information
- object classes
- image structure
- object categories
- keypoints
- depth map
- image representation
- multiscale
- feature extraction
- image sequences