Workshop on Deep Learning for Autonomous Driving (DLAD 2017)

Co-located with the

20th International Conference on Intelligent Transportation Systems (ITSC 2017)

Yokohama, Japan, October 16-19, 2017

Intelligent Transportation Systems (ITS) is the IEEE society for automotive broadly including autonomous driving. The Intelligent Transportation Systems Conference (ITSC) is the annual flagship conference of the IEEE Intelligent Transportation Systems Society and this year it is the 20th anniversary of the conference this year.

Deep learning has been progressing rapidly and disseminated via conferences like NIPS and CVPR. This workshop is an attempt to bridge the gap between latest research in Deep Learning and application to autonomous driving which is an active area of research in both academia and industry. Compared to other successful applications of deep learning, autonomous driving has its own set of challenges to achieve high levels of accuracy and reliability needed for a safety system. The first success of Deep Learning was mainly in visual perception via CNNs which has enabled applications like semantic segmentation which wasn't deemed possible before and expanding into classical geometric vision problems like Optical Flow and Structure from Motion. The other application areas like motion planning, sensor fusion, etc are in early stages of research. There is also the ambitious side of solving autonomous driving by a single deep learning model (end-to-end learning) and its variant of modular end-to-end with auxiliary losses for semantics. From a deployment perspective, processing power is still a bottleneck and there is steady increase of computational power where next generation platforms are targeting 10-100 TOPS.

This workshop aims to bring together the latest research of deep learning applied to various problems in Autonomous Driving. We are soliciting contributions in (but not limited to) the following topics:

    • Deep Learning for visual object detection (Object detection & Semantic Segmentation)

    • Deep Learning for geometric vision (Optical Flow, Structure from Motion & SLAM)

    • Feature Extraction and perception for sparse sensors (LIDAR & RADAR)

    • Deep Reinforcement learning for Motion Planning

    • Deep learning based sensor fusion and target tracking

    • Real-time implementation on embedded platforms (Efficient design & hardware accelerators)

    • Challenges of deployment in a commercial system (Functional safety & High accuracy)

    • Design of datasets (Synthetic datasets & Transfer learning).

    • Advanced ground truthing methods for training deep learning algorithms (simulation, automatic, semi-automatic)

    • End to end learning of driving (Single model & modular end-to-end)

    • Advanced topics on Deep Learning (Meta-learning, Multi-task learning)

    • Other Machine Learning algorithms and their interaction with Deep Learning (PGMs & Probabilistic Programming)