Video and Image Processing Cup
This competition is sponsored by the IEEE Signal Processing Society.
Traffic Sign Detection under Challenging Conditions
Dataset is available HERE
To officially attend the VIP cup, you should sign up from HERE
(Full detailed description is available HERE)
- May 16, 2017: We would like to encourage the competing teams to take behind-the-scenes pictures while working on the competition. You will be the first teams to work on VIP Cup so let’s make it special and show others later on how the competition process is thanks to your pictures. We will ask for these pictures as part of your submission package. We are confident that you will do a good job in the algorithmic side so you can also start thinking about demonstration ideas. The best three teams will give a presentation at ICIP 2017, which should include a demonstration/video showcase. Presentation session will be the time for you to shine and get all the recognition you deserve.
- May 15, 2017: There was misnaming in 6 of the video files. The challenge levels for the video sequences that start with 02_07_01_09 were changed from 00-04 to 01-05. And there was an extra underscore in 02_11_01_01_05_, which was removed.
- April 8, 2017: The detection task in VIP Cup 2017 includes localization and identification of traffic signs. As stated in the long description Section 2.4 and Section 2.7, the submission requires estimated detection files which should include sign types and coordinates of the bounding boxes for detected signs.
- March 30, 2017: Please do not forget to official signup to the VIP Cup from HERE!
- March 29, 2017: You can access the links for Google Drive, Microsoft OneDrive, and Dropbox from the Google form above.
- March 28, 2017: In case of Dropbox access issues because of heavy data traffic, you can utilize the OneDrive link. If you have already signed up to the Google form before and did not receive an email about the access links, you can refill the sign up form above and save the links.
1. Competition Overview
Robust and reliable traffic sign detection is necessary to bring autonomous vehicles onto our roads. State of the art traffic sign detection algorithms in the literature successfully perform the task over existing databases that mostly lack realistic road conditions. This competition focuses on detecting such traffic signs under challenging conditions.
To facilitate such task and competition, we introduce a novel video dataset that contains a variety of road conditions. In such video sequences, we vary the type and the level of the challenging conditions including a range of lighting conditions, blur, haze, rain and snow levels. The goal of this challenge is to implement traffic sign detection algorithms that can robustly perform under such challenging environmental conditions.
2. Competition Organization
The competition is open to any team that is eligible to participate and make a submission, which is due July 1, 2017 and the finalists of the competition are announced by August 1, 2017. Finalists will be selected by the organizers of the competition and will be invited to ICIP 2017 based on the overall performance whose details will be announced when the database is released on March 15, 2017. A judging panel will evaluate the three finalists to determine the ultimate winners at ICIP 2017, which will be held September 17-20, 2017. In addition to algorithmic performances, demonstration and presentation performances will also affect the final ranking. Presentations will be followed by questions of judging committee and general audience.
3. Organizing Team
The competition is sponsored by the IEEE SPS Multimedia and Signal Processing (MMSP) Technical Committee in collaboration wth the following organizing team members:
– Ghassan AlRegib
– Can Temel
– Tariq Alshawi
– Steve Chen
* VIP Cup is overseen by the Student Service Committee of the IEEE Signal Processing Society (Patrizio Campisi, Chair).