Notice

Remaining datasets for WAD Challenges are available for download at http://bdd-îdata.berkeley.edu/. Submission details will be available here in a couple of weeks, if not sooner.

The Video Segmentation Challenge is now live. We are excited for to host this competition, which provides a challenging data set of 800K street-view images with per-pixel semantic annotation, courtesy of Baidu's ApolloScape project. In this challenge, you are asked to segment out each instance of different objects in the test video. The video frames exhibit challenging cases of large lighting variations, heavy occlusions, and busy background. To get the best results, you are encouraged to explore the continuously annotated video frames that are only available in this data set.

Please feel free to ask your questions in this thread. Good luck!

Introduction

The CVPR 2018 Workshop on Autonomous Driving (WAD) is the combined venue for The 9th international Workshop on Computer Vision in Vehicle Technology (CVVT) and perception challenges with newly collected and fine-annotated large scale datasets. It aims to get together researchers and engineers from academia and industries to discuss computer vision applications in autonomous driving. In this one and half day work, we will have regular paper presentations, invited speakers, panel discussions, and technical benchmark challenges to present the current state of the art, as well as the limitations and future directions for computer vision in autonomous driving, arguably the most promising application of computer vision and AI in general.

PAPER Track

We invite the submission of original research contributions in computer vision addressed to:

· Autonomous navigation and exploration based on vision and 3D.

· Vision based driving assistance, driver monitoring and advanced interfaces.

· Vision systems for unmanned aerial vehicles.

· Deep Learning, machine learning, mathematical imaging and image analysis techniques in vehicle technology.

· Non-verbal and graphical information for remote-driver assistance of long-distance exploration.

· Performance evaluation without ground truth and reconstruction from one time measurements in natural environments.

· On-board calibration of multi-camera acquisition systems (stereo rig, multimodal, networks).

· Reconstruction without classical features such as planes, lines and linear objects and terrain generation from multi-view and omnidirectional camera networks.

· Large-scale computer vision and geo-localization for driving, navigation and exploration.

For details, please check WAD2018 Call for Papers

Submission details can be found at http://www.wad.ai/paper.html

CHALLENGE Track

We will host a challenge to understand the current status of computer vision algorithms in solving the environmental perception problems for autonomous driving. We have prepared a number of large scale datasets with fine annotation, collected and annotated by Berkeley Deep Driving Consortium or Baidu Inc. Based on the datasets, we have define a set of four realistic problems and encourage new algorithms and pipelines to be invented for autonomous driving. More specifically, they are

(1) Drivable Area Segmentation

(2) Road Object Detection

(3) Domain Adaptation of Semantic Segmentation

(4) Instance-level video moving object segmentation

Participation details can be found at http://wad.ai/challenge.html

Schedule
Papers
Challenge
Submission
March 27th
Launch
April 4th
Notification
April 10th
Close
June 11th
Camera-ready
April 12th
Paper Submission
June 11th
Camera Ready
June 18th
CVPR Workshop: June 18th, 2018
Workshop on Autonomous Driving : June 22th, 2018
Sponsor
News
2018.02.16
Web page is live.
2018.03.20
Paper submission deadline is now extended to March 27.
2018.04.04
The Video Segmentation Challenge is now live!