Submit your work any time from April 23rd, 2020 to June 8th, 2020.
A-list test set release date: May 6th
B-list test set release date: May 28th
Submissions deadline: June 8th
To register, email the following information and confidentiality statement to firstname.lastname@example.org.
Team name, Team members(Complete names), Organization name(Or Individual), Contact information, Include the following statement: ‘I agree to keep the data confidential’
Your team will be provided with 1000 short videos of the manually labeled training set, 200 short videos of the verification set, and 200 short videos of the test set. Videos are on average ten seconds long and based on one frame per second labeling.
Following registration, your team will download the project data to the local debugging algorithm and train the face parsing model to automatically perform face parsing segmentation according to each frame of the input video.
The competition team will release the A-list test set on May 6th. Participating teams will use the trained face analysis model to predict video frame images in the test set, generate a PNG gray-scale image, and then submit the data. List A will be updated daily and the highest ranking results from this stage will be posted. The test set for list B will be released on May 28th, and the final ranking of the competition will be based on list B. Note: the A-list test set is part of the B-list test set.
This competition aims to facilitate academic and creative exchange and to promote the development of graphic technology in China. The source videos used in this contest are collected from publicly available websites in line with relevant copyright rules and should be used for the purposes of this contest only.
Developers can access the link below for free training resources:https://www.nvidia.cn/developer/online-training
University challengers can be invited to participate in the University Ambassador Program of NVIDIA Deep Learning Institute (DLI) , or download AI Teaching Material , to enjoy free access to practical training resources of AI application and development by DLI.
Wechat account name for contact: DLIChina.
Learn more about NVIDIA Deep Learning Institute ( DLI ) and its University Ambassador Pro- gram, please visit : https://www.nvidia.cn/deep-learning-ai/education/?activetab=certification- tabs-3
Update time of competition results: 2020-06-14