DNNs have shown remarkable success in many computer vision and machine learning tasks. Despite a wide range of impressive results, current DNN based methods typically depend on massive amounts of accurately annotated training data to achieve high performance. DNNs lack the ability of learning from limited exemplars and fast generalizing to new tasks.
The Visual Intelligence Group in School of Data Science, Fudan University will hold the 1st Visual Intelligence Seminar on Few-shot Learning on January 29, 2021. We invite several distinguished speakers to share the recent progress on few-shot learning.
8:00 -- 8:10, Opening.
8:10 -- 9:10, Learning to Learn More with Less.
Speaker: Yuxiong Wang.
Host: Yanwei Fu.
9:10 -- 10:10: Meta-Learning: Representations and Objectives.
Speaker: Timothy M. Hospedales.
Host: Yanwei Fu.
10:10 -- 10:30, Break and Panel Discussion of Speakers in the Morning Session.
10:30 -- 11:30, 自监督学习进展、有效性初探及展望.
Speaker: Songcan Chen.
Host: Xiangyang Xue.
11:30 -- 12:10, Towards Good Practices in Self-Supervised Representation Learning.
Speaker: Yi Zhu.
Host: Li Zhang.
13:30 -- 14:30, New Meta-Learning Paradigms for Few-Shot Learning.
Speaker: Zhiwu Lu.
Host: Yanwei Fu.
14:30 -- 15:10, Learning universal representation for scene understanding.
Speaker: Li Zhang.
Host: Xiangyang Xue.
15:10 -- 15:50, Unsupervised Learning of 3D Objects from Images.
Speaker: Shangzhe Wu.
Host: Li Zhang.
15:50 -- 16:30, Self-supervised Representation Learning from Videos.
Speaker: Weidi Xie.
Host: Li Zhang.
16:30 -- 16:40, Break and Panel Discussion of Speakers in the Afternoon Session.
16:40 -- 17:20, Video Understanding In the Wild with Incomplete Supervision.
Speaker: Anurag Arnab.
Host: Li Zhang.
17:20 -- 18:20, Learning with few labeled data.
Speaker: Tao Xiang.
Host: Yanwei Fu.
18:20 -- 18:40, Panel Discussion and Finish.
| |
|
|
|
|
|
|
|
|
|
|
|
For more details, please check the seminar website:
http://www.sdspeople.fudan.edu.cn/fuyanwei/VI_seminar/fudan-vi-seminar.github.io/index.html
