Semantic Segmentation of Small Objects and Modeling of Uncertainty in Urban Remote Sensing Images Using Deep Convolutional Neural Networks.
Michael KampffmeyerArnt-Børre SalbergRobert JenssenPublished in: CVPR Workshops (2016)
Keyphrases
- remote sensing images
- semantic segmentation
- street scenes
- convolutional neural networks
- remote sensing
- multispectral
- change detection
- object classes
- satellite images
- object categories
- scene classification
- conditional random fields
- weakly supervised
- superpixels
- pascal voc
- object class
- hyperspectral
- bounding box
- object detection
- data sets
- d objects
- multiple objects
- object segmentation
- image fusion
- pattern recognition
- machine learning