CI-Net: a joint depth estimation and semantic segmentation network using contextual information.
Tianxiao GaoWu WeiZhongbin CaiZhun FanSheng Quan XieXinmei WangQiuda YuPublished in: Appl. Intell. (2022)
Keyphrases
- contextual information
- semantic segmentation
- depth estimation
- context aware
- stereo vision
- depth map
- stereo matching
- depth information
- superpixels
- high level
- dynamic scenes
- conditional random fields
- object categories
- weakly supervised
- stereo images
- disparity map
- higher order
- scene understanding
- real scenes
- viewpoint
- machine learning