When AI+computer vision meets various industries

Zhang Lei: Computer vision is the core problem of AI, because 70% of the information in people's various perception abilities comes from our visual system. Computer vision already has a lot of landing technologies, such as face recognition, and we will explore more scenarios today. We are very fortunate to invite researchers and scholars with both research background and practical experience.

Xu Yihua: I am the earliest entrepreneur among the members of Microsoft Research Asia, and I paid attention to two scenarios at the beginning of my business: one is the industrial field; The second is the medical field. The common point between these two fields is that the scene is absolutely controllable, the light source is controllable in all aspects, and the image in these controllable scenes is relatively simple, and the talent of our research algorithm can play a value. I have been in this industry since the end of 2005. Fortunately, in 2019, we became one of the three companies to meet the science and technology innovation board for the first time, and the meeting also represented the country's recognition of us who are engaged in hard science and technology engineering.

Zhang Lei: Dr. Yihua's experience perfectly illustrates the "concentration" and "patience" that Dr. Wang Jian (mentioned in the "Roundtable Forum: Academician Dialogue") talked about, without which it is impossible to achieve the results that we have today.

Zeng Wenjun: Then I am verifiable and the first to come out of Microsoft Research Asia to found the school. After leaving Microsoft Research Asia, he joined Eastern University of Technology, which was founded by a 300 billion donation from the famous entrepreneur Mr. Yu Renrong, and is a key project supported by the Zhejiang Provincial and Ningbo Municipal Government. The goal is to host a private, new type of innovative university. With the help of the world's advanced experience in terms of school philosophy, mechanism, culture, training goals, etc., we will make some cutting-edge attempts and explorations, and also hope to solve some problems for the country in the education industry. The school is currently being organized and hopes to receive more attention and support.

Hua Gang: Convenience Bee is a convenience retail enterprise founded by Zhuang Chenchao and his partners in 2016, and today, after 5 years, we can call ourselves the largest company directly operated convenience chain stores, with more than 2000,<> stores nationwide, and sinking new stores to second- and third-tier cities at a rate of hundreds every month. The reason for the massive expansion of our company's direct store model is because the entire operating system is driven by numbers, algorithms and systems. I believe that the biggest contribution of computer vision to the industry is the system formed through computer vision, which can digitize the information in the offline physical world. At the same time, the digitized information can eventually fit (adapt) to the decision-making system. Convenience Bee is such a new retail company.

Zhang Lei: We see that computer vision is beginning to enter more and more specific application scenarios, such as industrial vision, new retail, and other fields. In the past decade, computer vision has made a very big breakthrough, if you take the current technology to ten years ago, it was all black technology. At the same time, we are starting to see that technology has begun to face more landing problems, and we also have a lot of very good researchers entering the enterprise to look at specific problems. Today, I would like several guests to share what kind of problems they have encountered in the industry they are engaged in? What are the characteristics of the problem scenario?