MultiNet: Real-time Joint Semantic Reasoning for Autonomous Driving
Accepted version
Peer-reviewed
Repository URI
Repository DOI
Change log
Authors
Abstract
While most approaches to semantic reasoning have fo- cused on improving performance, in this paper we argue that computational times are very important in order to en- able real time applications such as autonomous driving. To- wards this goal, we present an approach to joint classifi- cation, detection and semantic segmentation via a unified architecture where the encoder is shared amongst the three tasks. Our approach is very simple, can be trained end-to- end and performs extremely well in the challenging KITTI dataset, outperforming the state-of-the-art in the road seg- mentation task. Our approach is also very efficient, allow- ing us to perform inference at more then 23 frames per sec- ond. Training scripts and trained weights to reproduce our results can be found here: https://github.com/ MarvinTeichmann/MultiNet