I am a PhD student at Stanford working on computer vision with Prof. Fei-Fei Li, focusing on large-scale recognition.
- Thesis title: "Designing and Overcoming Challenges in Large-Scale Object Detection"
- Details: "Designing challenges" refers to my work on the object detection task of the ImageNet Large Scale Visual Recognition Challenge and to the related computer vision (ICCV13, Corr14, CVPR15) and human-computer interaction (CHI14) contributions. "Overcoming challenges" refers to my work on developing novel object detection algorithms for weakly supervised localization (ECCV12), faster object detection (CVPR10), and others.
- Expected graduation: August 2015
- March 2015: Our ImageNet challenge paper is accepted to IJCV.
- March 2015: Two papers accepted to CVPR2015:
- Best of both worlds: human-machine collaboration for object annotation (with Li-Jia Li and Li Fei-Fei)
- Joint calibration of Ensemble of Exemplar SVMs (with Davide Modolo, Alexander Vezhnevets and Vittorio Ferrari)
- January 2015: Fei-Fei Li, Rick Sommer and I are co-directing the first-ever Stanford AI Lab Outreach Summer program. It is a two-week full-time summer program for high-school students. Our goal is to encourage interest in the field of AI among underrepresented minorities.
- January 2015: I'll be the Publicity and Press chair for CVPR2016.
- December 2014: I'm co-organizing two workshops and a tutorial at CVPR2015:
- BigVision: International Workshop on Large-Scale Visual Recognition and Retrieval with Jason Corso, Jia Deng and Yuanqing Lin
- WiCV: Women in Computer Vision workshop with Judy Hoffman, Adriana Kovashka, Brigit Schroeder and Ning Zhang
- ImageNet Large Scale Visual Recognition Challenge tutorial with Jonathan Krause, Karen Simonyan, Yangqing Jia
- September 2014: This year's ImageNet challenge and our paper are featured in the New York Times, MIT Technology Review, CBC radio, Neuen Zürcher Zeitung, NewScientist.
- September 2014: Our new paper which describes the collection of the ImageNet Large Scale Visual Recognition Challenge dataset, analyzes the results of the past five years of the challenge, and even compares current computer accuracy with human accuracy is now available.