According to the conference report of CVPR2020, TAL’s Academy of Engineering team stands out from the 69 teams in the World Champion of the EmotioNet Challenge, a global top academic competition. Specifically, the team proposed a solution to the inconsistent overfitting problem of multi-task learning in facial expression recognition, which improves the accuracy and F-value of the model. Currently, this technology has been widely used in the teaching business of TAL and is shared with the whole industry through TAL AI’s open platform.

CVPR (International Conference on Computer Vision and Pattern Recognition), known as the "Olympic Games" in the field of computer vision, will be held in Seattle, USA in June 2020.

EmotioNet Challenge, an important part of this conference, is one of the most prestigious  international academic competitions in the field of facial expression recognition, and its research results are published in CVPR and ICCV (International Computer Vision Conference), two of the three top conferences in computer vision. 

EmotioNet has attracted a large number of prestigious international universities and corporate teams to participate.EmotioNet 2020 is organized by CVPR organization and the Ohio State University. Many professional teams including TAL, Alibaba, University of Science and Technology of China, Chinese Academy of Sciences, Magdeburg University, Nanyang Technological University participated in this competition. 

For a long time, the difficulty of landing AI technology is a common issue in the industry for its application requires the support of massive data and rich scenarios. In the past ten years, TAL has accumulated a large amount of data resources and garnered valuable experiences in the integration of AI and Education. In recent years, AI education in China is witnessing a rapid growth. As a leading enterprise in the education industry, TAL is actively taking the responsibility of promoting AI technology and working with industry partners to address key issues related to the applications of smart education, as well as the education industry. Ultimately, TAL's goal is to build a vibrant and creative education ecosystem.