[Promotion] Research team led by Professor Usaimon of Software Convergence University, and Minister of Science and ICT Award for Artificial Intelligence Grand Challenge
- SKKU National Program of Excellence in Software
- Hit2038
- 2022-01-14
Research team led by Professor Usaimon of Software Convergence University, and Minister of Science and ICT Award for Artificial Intelligence Grand Challenge
[Photo] Professor Usaimon of Software Convergence University. Research team.
The Data-Based Convergence Security Laboratory (DASH LAB, advisor Usimon Sungil, master's course in software/convergence security track, Lee Han-bin, Kim Jung-ho, and Kim Jin-beom) won the first prize in the fourth phase of the 2021 Artificial Intelligence Grand Challenge Track 1 (behavioral recognition) for the second consecutive year.
The Artificial Intelligence Grand Challenge is a challenging and competitive R&D competition in which participants compete with algorithms developed to solve the presented problems. The fourth phase of the third competition was held online from November 12 to the 14th of last year under the theme of "Use artificial intelligence technology to detect people in need quickly and accurately." In the first and second stages of competitions held in 2020, he ranked first and third, respectively, and in this competition, he also ranked first and won the Minister of Science and ICT Award.
This competition aims to detect people who show abnormal behavior quickly and accurately in limited computing power. In order to overcome the trade-off of speed and performance, this laboratory used a strong post-processing algorithm to enable real-time processing even with large models. In addition, the trained model used the VFP290K dataset recently announced in NeurIPS 2021 Datasets & Benchmarks Track (VFP290K: A Large-Scale Benchmark Dataset for Vision-based Fallen Person Detection | OpenReview) showed excellent performance in this challenge.
Meanwhile, DASH Laboratory continues to produce excellent research results by publishing papers in AAAI, ICML, KDD, NeuralPS, WWW, and CIKM, the world's most prestigious artificial intelligence societies.
Source: http://www.aitimes.com/news/articleView.html?idxno=141936
https://www.yna.co.kr/view/AKR20211208151200017?input=1195m