How big is the 1 mm²? Ballpoint pen tip size. If such a small point is placed on a cloth running at a high speed of tens of meters per minute, the difficulty of detecting it can be imagined. For non-woven masks as raw materials, can not tolerate a little flaw, flaw detection has become a problem in the industry. Today, the Artificial Intelligence Engineering Application Laboratory of Zhengzhou University has innovated the application of non-woven fabric quality inspection based on the deep learning platform of flying oars, bringing new breakthroughs.

After nearly a year of research and development, the laboratory has successfully developed a customized and lower-cost intelligent classification system for non-woven fabric defect detection by using paddles in conjunction with Jiangsu Hemi Intelligent Technology Co., Ltd. (hereinafter referred to as "Hemi Intelligent"), which further optimizes the way the originally expensive detection equipment loads stylized traditional algorithms, and is still iteratively updating, allowing artificial intelligence technology to enter the industry from the laboratory.

Atletico Atletico, director of the Artificial Intelligence Engineering Application Laboratory at Zhengzhou University, said, "Since the establishment of the laboratory, we have been looking for practical opportunities for students to meet the needs of enterprises. Through this industry-research collaboration, students have given full play to the learning accumulation of AI professional courses, and have deeply realized the difference between theoretical knowledge and technology landing in practice. The ability of technological achievements to be effective in industrial applications has also greatly encouraged teachers and students."

Zheng Da and Flying Paddle School and Enterprise Build Together to Cultivate Compound AI Talents

At the WAVE SUMMIT Deep Learning Developer Summit in 2021, Zhengzhou University and Baidu Feizhou signed a contract to build the Artificial Intelligence Engineering Application Laboratory of Zhengzhou University. This is one of the first three innovation and entrepreneurship laboratories signed by Baidu Feizhou and universities. Different from the regular path of Baidu's cooperation with the College of Computer Science in colleges and universities, the laboratory is an artificial intelligence engineering application teaching and training platform jointly built on the basis of the work of the artificial intelligence laboratory of the School of Mechanics and Safety Engineering of Zhengzhou University. the aim is to enable the training of "artificial Intelligence X" new engineering talents.

In the words of Atletico Madrid, director of the Artificial Intelligence Engineering Application Laboratory of Zhengzhou University, "We don't produce artificial intelligence algorithms, we are porters of artificial intelligence algorithms." He believes that letting people with a foundation in mechanics learn and apply artificial intelligence is a shortcut to use artificial intelligence technology to solve problems in the engineering field, because mechanics is the foundation of engineering science, and its tentacles can penetrate into many engineering fields. And let artificial intelligence in more industrial scenarios in the ground application, benefit the production line, is also Baidu flying oar construction and development ideas, the two sides hit it off.

Previously, the artificial intelligence laboratory of the Institute of Mechanics of Zhengzhou University used a tutorial system to lead students to do science and technology. But at that time, teachers and students did not systematically learn and master artificial intelligence technology. Until after the cooperation with the paddle, the laboratory was upgraded, and the teacher Wei Hongtao of the Institute of Mechanics participated in the in-depth study teacher training organized by Baidu Paddle, began to systematically master the professional knowledge of artificial intelligence and carry out practical exercises, and then guided the students to do research and development according to the real needs of enterprise intelligence.

The "AI Engineering" team of the laboratory has continuously improved its technical level, and has repeatedly won awards in many competitions such as "China Robot Competition", "artificial Intelligence Creative Competition", "Mechanical Design Innovation Competition", "Software Cup", "Challenge Cup" and so on.

In-depth industry first-line practice, non-woven defect classification detection upgrade

Wei Hongtao introduced that there were 7 project members participating in the research and development of the non-woven intelligent quality inspection system. The students all went deep into the production line for on-site investigation, development, and testing, and exercised their ability to solve problems in actual combat.