The National University of Singapore (NUS) Learning and Vision (LV) team clinched the top first place awards in the ICCV 2017 MS-Celeb-1M Large-Scale Face Recognition Hard Set / Random Set / Low-Shot Learning Challenges. The award ceremony was held in Venice on 28 October 2017.
The project was a collaboration between the NUS ECE Department’s LV group (Dr. Feng Jiashi, Mr. Zhao Jian, Mr. Wang Zhecan) and Panasonic R&D Center Singapore.
MS-Celeb-1M is the well-known “World Cup” challenge in the field of face recognition. It is organized by Microsoft Research, established on June 29, 2016. The 2016 Challenge focused on the challenge of recognizing one million celebrities in the real world, and it was organized together with ACM Multimedia Conference 2016 (ACM MM 2016).
The 2017 Challenge which was organized together with the International Conference on Computer Vision 2017 (ICCV 2017), include two main challenges:
- Recognizing One Million Celebrities, officially evaluated on Hard set and Random set, respectively;
- Low-shot Learning.
The 1st challenge was to recognize one million celebrities from their face images and identify them by linking to the unique entity keys in a knowledge base. This task introduces very large scale face recognition with disambiguation, while the 2nd challenge was to investigate the problem of low-shot face recognition, with the goal to build a large-scale face recognizer capable of recognizing a substantial number of individuals with high precision and high recall.
