Low-Power Neural Networks for Semantic Segmentation of Satellite Images.
Gaetan BahlLionel DanielMatthieu MorettiFlorent LafargePublished in: ICCV Workshops (2019)
Keyphrases
- low power
- satellite images
- semantic segmentation
- neural network
- power consumption
- high speed
- low cost
- multispectral
- conditional random fields
- superpixels
- remote sensing
- change detection
- scene classification
- high resolution
- weakly supervised
- pattern recognition
- pascal voc
- object categories
- object class
- object detection
- image understanding
- image analysis
- object recognition
- real time
- object segmentation
- input image
- cmos technology
- ultra low power
- object classes
- long range
- multi view
- graphical models
- multiscale