FASSD-Net: Fast and Accurate Real-Time Semantic Segmentation for Embedded Systems.
Leonel Rosas-AriasGibran Benitez-GarciaJosé Portillo-PortilloJesus Olivares-MercadoGabriel Sanchez-PerezKeiji YanaiPublished in: IEEE Trans. Intell. Transp. Syst. (2022)
Keyphrases
- signal processing
- embedded systems
- semantic segmentation
- real time systems
- low cost
- real time
- resource limited
- street scenes
- image processing
- superpixels
- conditional random fields
- label transfer
- computer vision
- field programmable gate array
- software systems
- scene classification
- weakly supervised
- image understanding
- object classes
- multiscale
- image set
- object class
- open source
- active learning
- machine learning