A Top AI event in China: How does a rock make waves?

The paper contains 3419 words and is expected to last 7 minutes

Creativity always comes from inspiration, yu Yongda firmly believes this.

“One day WHEN I was sleeping at my desk, I heard someone at the next table writing different notes with different sounds. And I thought, maybe I can tell what’s written by reading voices. Later, I saw that when the teacher was lecturing, because the screen in the classroom did not have the touch screen function, he had to go to the computer to switch each PPT page. If we can make screens, blackboards and even walls look like touch screens, we can control smart devices anytime and anywhere.”

No sooner said than done. It took Yu and his classmates four months to turn the idea into reality — the deep learning-based smart home control system named “Finger dance” can track sound signals to recognize gestures.

The end of November in Zhejiang Deqing has not entered the strict sense of winter. In deqing International Convention Center, Yu Yongda and his classmates from Dalian University of Technology showed off their “finger dance” here and won the special prize of the Enabling Group of the 2019 China University Computer Contest – Artificial Intelligence Creativity competition (C4-AI).

Yu yongda is one of many young people who have turned their ideas into reality. These ideas are like stones thrown into a still water, creating ripples, and there is still a long way to go before they can make waves.

An event with a mere 2.6 per cent entry rate

The AI Zhifubao developed by He Xiaoyu and his team at Central South University has gone one step further and is already operational. The skin disease classification system developed by using a large number of real cases of skin disease patients in Xiangya Hospital of Central South University can diagnose whether the person in the photo is suffering from skin disease online by uploading a photo.

“We have been working with Medcom for two years and are industry leading in terms of data, innovative algorithms and paper results.” Xie Bin, a teacher from Central South University who led the team to participate in the competition, is full of confidence in this project. In the end, it won the special prize in the CREATIVE team of the C4-AI Competition.

To solve the problem of pest control, students at Hainan University developed a pest control robot based on Beidou navigation and thermal imaging vision technology. In order to better identify potential depressed patients on social media platforms, students at Zhejiang University of Technology built an emotion classification and discrimination model. In order to make the operation of college canteen more intelligent and efficient, Zhejiang Sci-Tech University through image recognition and face recognition, really achieve the “face to eat”…… It is not easy for the creators of these events to stand on the podium. From 1675 teams registered in April this year to 43 teams participated in the national finals in November, only 2.6% of the participation rate, the gold content of the competition is self-evident.

At the C4-AI award ceremony, Pan Yunhe, academician of The Chinese Academy of Engineering (second from right), Wu Xiaozheng, Deputy Party Secretary of Zhejiang University (first from left), and Li Xuanyya, director of Baidu Campus Brand Department (first from right) presented awards to the special prize team.

Xue Jianru, a professor at the Institute of Artificial Intelligence and Robotics at Xi ‘an Jiaotong University, has been serving as a judge since c4-AI was born in 2018. “Compared to last year, this year’s project is much better in terms of both technology and system integrity, and hardware presentation.” Chen Linqiang, professor of Computer School of Hangzhou Dianzi University, believes that the biggest highlight of the competition is that “enterprises deeply participate in the organization and planning of the competition, so that the projects of colleges and universities can be better implemented”.

A new era of AI education was opened by the competition

2019 is the first year of artificial intelligence talent training, which has become a consensus in the industry.

It goes without saying how important AI is to the country’s development, and talent cultivation is a top priority. Zhou Ming, deputy director of the Education Examination Center of the Ministry of Industry and Information Technology, revealed in 2018 that China has a shortage of AI talent of more than 5 million people. However, the current shortcomings are also obvious: with the traditional talent training mode, it is difficult to get the talents that enterprises, society and even the country really need. Lack of professional teachers in colleges and universities, lack of authoritative teaching materials in classrooms, outdated theories divorced from reality, and difficult to implement scientific research achievements… Many pain points need to be addressed.

So, how should AI talents, especially those majoring in AI+X, be cultivated? The government, universities and even enterprises are exploring. The prestigious AI-themed events have found a delicate balance, bridging the divide between campus and market, theory and practice.

There is no better proof of this conclusion than the mushrooming of competitions. The Chinese University computer contest – Artificial Intelligence Creative Competition is one of them.

“A work from creativity to system construction, and then to show it clearly to the outside world, the test is students’ ability to find problems, analyze problems and solve problems, as well as hands-on ability, expression ability. After all, we want to cultivate people with comprehensive abilities.” Prof Xue says the competition is an effective way to test innovation, like a baton telling people in which direction to work.

This is why Xie Bin led students to participate in the competition for two consecutive years.

“Through the competition, students can find the direction they are interested in, understand the needs of the development of the industry, and have the opportunity to connect with leading enterprises in the industry. It’s a great development model for students. After we won the first prize last year, our team leader Song Di joined Baidu and became a product manager. Companies have also selected talented young people in this way.

The competition is like a high-gain, high-power amplifier, amplifying the ripples of the competitors’ ideas until they form an unstoppable wave.

Why is the ecology of talent cultivation so important?

Earlier, Dr. Wang Haifeng, Baidu’s chief technology officer, said at the award ceremony marking the 15th anniversary of the Baidu Star Competition that ARTIFICIAL intelligence, as the core of the fourth industrial revolution, is increasingly integrated with all industries. Baidu flying-paddle deep learning platform and EasyDL customized training service platform provide support for the competition. We hope young students can complete a variety of intelligent ideas with Baidu AI technology, jointly promote the development of AI industry in the whole society, and get self-inspiration and growth from it.

This is especially true for AI education. Relying on the traditional talent training mode, we can never cultivate talents who can adapt to the industrial development and keep pace with The Times. The best solution is to establish an ecological closed loop in which talent cultivation and industry development promote each other.

Universities gather high-quality algorithms and industries accumulate a large amount of data. However, the separation of market and scientific research is bound to make the two move towards split development. A good ecological closed loop will break down all these barriers through appropriate channels and platforms, formulate reasonable policies and regulations, cultivate high-quality teachers, compile advanced course materials, implement projects with strong implementation, and eventually cultivate usable and easy talents.

In this ecological closed loop, the government, universities and enterprises are indispensable.

In addition to the Ministry of Education, Zhejiang University, Deqing County, this competition also has a corporate figure – Baidu Inc. In addition to recruiting a large number of participating teams at home and abroad by using the resources of universities that have been deeply cultivated for many years, Baidu also provides the deep learning platform of flying OARS and its customized training service platform EasyDL as the technical platform of the competition. In order to fully tap the potential of participants and further utilize the technology platform to improve the project, Baidu also conducted a four-day training course for 107 finalists in August.

As the first and currently the only open source deep learning platform in China, it is inevitable that flying OARS will be used to support this national competition. “There are many good platforms in the world, but in the context of global competition in AI, it is necessary for us to advocate and develop domestic open source platforms to better promote the development of open source ecology. As far as I know, the performance of our domestic platform is actually better in some aspects than that of foreign open source platforms such as Google.” Professor Xue Jianru has been thinking deeply about platforms.

Focus on AI education, not just AI education

Traditionally, talent development has always been the business of the government and education circles, and there seems to be no good reason for enterprises to get involved. But in fact, through the industry research found that Baidu had already started the layout of AI education ecology before the competition. It seems that Baidu’s “ambitions” go far beyond simply supporting an event.

The ultimate goal of all initiatives of an enterprise is to develop better in the industry. As a wild goose enterprise, Baidu is shouldering the special mission of making Chinese AI leading the world. In the development of AI education on the road of exploration, Baidu has made various attempts. Through the oar platform, Baidu has not only built a cooperative ecology of collaborative education, but also laid out its development strategy in all aspects of AI education in colleges and universities.

Apart from anything else, it seems that only an enterprise like Baidu can have hard and soft power to support the development of education in an industry.

Strategic cooperation with government, universities, independent research and development of open source platform, technical contests, teacher training, carry out online courses, writing professional teaching materials, etc., these measures not only promote the AI talent cultivation in China, from baidu itself, also promote the independent platform, mining the excellent talents, but also the layout strategy in the future… Professor Chen Songcan of Nanjing University of Aeronautics and Astronautics put it simply: “The most important thing for artificial intelligence is landing. Whoever grabs outstanding talents first will also grab future dividends.” In the process of “promoting learning with competition”, digging and absorbing more “fresh forces” like Song Di is probably the most precious wealth of Baidu.

The AI ecosystem baidu wants to build does not only stop at universities, a large number of industry developers have become a key group of Baidu AI enabling.

From the release of China’s first training standard for professional and technical personnel in the AI field to the launch of PaddleCamp, AI Fast Track, Whampoa College and other training programs for industry developers at different levels, Baidu’s “chess game” has gone beyond simple “teaching” and “learning”, but intends to lead the development of the whole industry. Even pushing countries to overtake in this area on corners. Up to now, Baidu’s flying-paddle deep learning platform has served more than 1.5 million developers, with more than 65,000 enterprise users and 169,000 models released on the customized training platform alone.

The integration of national development strategy and enterprise development trend of Baidu, how much energy can be released, how high waves, time will prove.

Leave a comment like follow

We share the dry goods of AI learning and development. Welcome to pay attention to the “core reading technology” of AI vertical we-media on the whole platform.

(Add wechat: DXSXBB, join readers’ circle and discuss the freshest artificial intelligence technology.)