[summary] some panels and videos seen

1. 20220615【AI for Science Physical information driven deep learning 】Panel
   Panel issue ：
   1. Compared with traditional numerical methods, neural networks have potential advantages in which problems ？
   2. Embed prior physical knowledge into deep learning models and algorithms , It can effectively improve learning efficiency and ability . What are the important developments and trends in this direction ？
   3. Neural networks are solving PDE What are the important theoretical issues ？
   4. The theory that physical information drives Neural Networks 、 What are the main challenges in algorithms and Applications ？ Is there a solution ？
   5. Visual data ( Especially video data ) Most of them are observations of the real physical world , How physical information can be applied to machine vision ( Such as 3D testing 、 Target tracking 、 Relationship modeling, etc ) To promote ？
2. 20220323【 I want to find you ：2D/3D Object detection and location 】Panel
   Panel issue ：
   1. What new inspiration will the scientific research progress in the field of natural language bring to computer vision ？
   2. Transformer What are the applications and research prospects in the field of object detection and location ？
   3. 2D/3D What are the new directions in visual understanding ？
   4. How to reduce detection Required supervision ？
   5. How to deal with detection Long tailed distribution in ？
   6. 2D/3D What are the difficulties worthy of further study in visual positioning ？
3. 20220427 Special report 【 Practical target tracking technology in complex scenes 】 Luhuchuan ( Dalian University of Technology )