TAL's AI Team Wins Championship Award at a Top International Academic Contest

News provided by

Apr 21, 2020, 21:17 ET

BEIJING, April 21, 2020 /PRNewswire/ -- The international Conference of Computer Vision and Pattern Recognition (CVPR2020), one of the top Artificial Intelligence academic conference, has recently revealed the grant challenge results in a series of categories. TAL's AI Lab has won the EmotioNet Challenge, an award for human facial expression recognition, surpassing 69 other teams from around the world.

Demonstrating China's AI competitiveness to world by winning the top-notch AI academic conference

Known as the Olympic Games of computer vision field, CVPR will be held in Seattle, Washington, USA in June 2020. The EmotioNet Challenge is one of the most revered academic competitions in the field of human facial expression recognition. The research methods used to tackle this challenge have been published on CVPR and ICCV (International Conference of Computer Vision), two of the top three computer vision conferences in the world. Over the years, the event has attracted a large number of teams from top universities and companies worldwide.

EmotioNet 2020, sponsored by CVPR and Ohio State University, has attracted teams from Alibaba, China University of Science and Technology, Chinese Academy of Sciences, Otto-von-Guericke-Universität Magdeburg, Nanyang Technological University, and other organizations. After fierce competition, TAL's AI Lab won the championship, demonstrating TAL's growing AI technology capabilities to the world.

In the competition, TAL's AI Lab has proposed a solution to the inconsistencies of overfitting during multitask learning in human facial expression recognition, which has significantly improved the accuracy and the F-score of the model. It is one of the many AI abilities that TAL has been able to develop and applied to multiple TAL products and services. Currently, this technology is also made available to the entire education industry through TAL's Open AI Platform.

Enlarging investments in the AI technology R&D for wider applications

Applying AI technology to the education products is often challenging due to the lack of data covering all forms of educational products and services.

Over the past decade, TAL has accumulated a massive amount of data and has continued exploring various education settings such as face-to-face (in-person), online one-to-one, online dual-teacher, and other methods of teaching. Understanding the changing demands with data generated from various types of education products, TAL is in a good position to further develop "AI + Education" technologies for students. The AI models improve as the types of TAL education products continue to grow.

TAL education group consistently steps up its efforts in research and engineering development. The AI Lab, the Brain Science Lab, the Silicon Valley R&D Center, the Big Data Center, and the Postdoctoral Workstation were all founded to better fulfill TAL's mission, "Making education better with love, science, and technology".

TAL keeps breaking new ground in the research and development of cutting-edge technologies. Last November, Hurun Research Institute listed its AI Intellectual Property in China 2019. A list comparing over 500 Chinese AI Companies based on their comprehensive development, innovation abilities, and technology maturity. TAL is ranked among the top 10. TAL also attended the AAAI Conference on Artificial Intelligence 2020, in February of this year. With the team's two latest research findings, TAL was invited to take the lead in organizing two symposiums on "smart education".

TAL has developed more than 100 AI capabilities for education in eight categories including graphics, voice, data mining, and natural language processing, etc. It has created more than 10 AI solutions based on different education scenarios, covering all aspects of teaching, learning, testing, exercises, and assessment. TAL has also developed many unique and/or leading abilities throughout the process of building its own AI platforms. This includes: containing infrastructure for storage, annotation, training, and online services.

Building a new diversified ecosystem for smart education in China

Education AI has been growing vigorously in recent years. As a leading company in the education sector, TAL is actively promoting the joint development of "AI+Education" and the advancement of education sector with its partners.

In August 2019, the Ministry of Science and Technology of China approved TAL to build the National Smart Education Platform for Open Innovation of Next Generation AI. This initiative means that TAL has officially become the "national team" behind exploring "Education+AI". The platform is expected to support the thorough integration of AI and education, letting smart education empower all stakeholders and improve teaching efficiency, management, and the opening-up of schools.

The National Smart Education Platform for Open Innovation of Next Generation AI will concentrate on learners, enhancing core links between education and education services step by step. By promoting supply-side reform of the education service sector, the platform strives to both increase productivity and cut the cost among education. Through breakthroughs, distribution of education resources, and realizing data-driven individualized teaching at scale, it will bring changes to educational concepts and models. Meanwhile, the R&D threshold for small to medium enterprises, as well as the risk in commercialization for education institutions, will be lowered. The enthusiasm for innovation and entrepreneurship in smart education will be unleashed.

TAL has technological partnerships with Tsinghua University, Institute of Computing Technology of Chinese Academy of Sciences, Beijing Normal University and Stanford University, jointly contributing to an open tech ecosystem ranging across both public and private institutions.

TAL Education Group expresses its aspiration to work with all stakeholders to meet learners' core needs, solving major issues in the application of smart education, and furthering the development of education industry. Focusing on broad AI technologies and platform-based services, TAL wishes to establish an AI open innovation platform for smart education that features independent and controllable basic software and hardware. An open exchange of general technology, data, effective application, and a vibrant innovation ecosystem; this will eventually create a new and diversified smart education ecosystem. A symbiotic environment that is made to cover all methods and scenarios within education.

SOURCE TAL Education Group