FreDSNet: Joint Monocular Depth and Semantic Segmentation with Fast Fourier Convolutions.
Bruno Berenguel-BaetaJesus Bermudez-CameoJosé Jesús GuerreroPublished in: CoRR (2022)
Keyphrases
- semantic segmentation
- street scenes
- label transfer
- conditional random fields
- superpixels
- scene classification
- weakly supervised
- depth estimation
- image understanding
- object class
- object categories
- depth information
- image sequences
- object classes
- pascal voc
- depth map
- multiscale
- viewpoint
- object recognition
- three dimensional
- long range
- image processing
- image representation