Figure 1: Leading Board of the “AI Meets Beauty” Challenge (Image from https://challenge2019.perfectcorp.com/#item11 )

ACM International Conference on Multimedia is the premier international conference in the area of multimedia within the field of computer science, and also the only A-Class conference certificated by the China Computer Federation (CCF) in the multimedia research field. This year, the conference will be held in Nice, France in October. Since 1993, ACM Multimedia has been bringing together researchers and practitioners from academia and industries to present innovative research results and discuss recent advancements.

As the revolution of artificial intelligence (AI) is underway around the globe, AI has already begun to reshape industries including the multibillion dollar beauty and personal care industry. Hence, a special part of the ACM conference —— the “AI Meets Beauty” Challenge, was held to explore the boundaries of computer science and art. The challenge focuses on the fine-grained recognition task of beauty and personal care products, on a provided large-scale image dataset of over half million images of beauty and personal care products, namely the Perfect-500K.

Shortly before the competition, the USTC-NELSLIP team also won the second place in the GrandChallenge of 106-p Facial Landmark Localization held by the renowned IEEE International Conference on Multimedia Expo (IEEE ICME).

IEEE ICME has been the flagship multimedia conference sponsored by IEEE ComputerSociety, CircuitsandSystemSociety, SignalProcessingSociety and CommunicationSociety since 2000. It serves as a forum to promote the exchange of the latest advances in multimedia technologies, systems, and applications from both the research and development perspectives of the circuits and systems, communications, computer, and signal processing communities. Specifically, the GrandChallenge of 106-p Facial Landmark Localization Competition is aimed to promote the development of research on face landmark localization, especially dealing with the complex situations, e.g. large face poses, extreme expressions and occlusions. 106-key-point landmarks were provided in the given dataset to enable abundant geometric information for face analysis tasks.

Figure 2: Leading Board of the GrandChallenge of 106-p Facial Landmark Localization Competition (Image from https://facial-landmarks-localization-challenge.github.io)

The team, consisting of Prof. YU Jun and graduate students XIE Haonian, XIE Guochen, LI Mengyan, achieved an AUC score of 82.68. The team was also the only university team among the top three winners.

(Written by ZHAO Shuxin, Edited by LU Hongyu, USTC news center)