Middle school students won the three-year Youth Innovation Award: similar to the patent of the same surname professor, the patent page has been deleted
Surging News Reporter Zhang Jing Zhang Wei Yang Yang Intern He Qingyi Li Antian Zhong Yixuan
After the 6th grade primary school student won the prize for cancer research but was disqualified for violating the "project research report must be written by the author himself", a reporter from the Surging News (www.thepaper.cn) survey found that there was a student in the National Youth Science and Technology Innovation Competition It has been awarded for three consecutive years, and the achievements have spanned physics, computer integration, and artificial intelligence.
One of the Duan students’ award-winning achievements is highly similar to a national patent filed earlier, and the inventor of the patent has the same name as the instructor of the student’s winning project.
In addition, the second inventor of the patent is Duan Shukai, and he is currently a professor, doctoral tutor and associate dean of the School of Artificial Intelligence of Southwest University. According to the official website of the National Youth Science and Technology Innovation Contest, Duan Duan’s three awards were in his second year of junior high school, third year of junior high school, and first year of high school (2017-2019).
The inventor of the patent has the same name as the instructor of the student’s award-winning project. The patent was applied on January 23, 2019. The application date of the 34th National Youth Science and Technology Innovation Contest is from March 20 to April 20, 2019.
Another award-winning achievement of the classmate, the “Unmanned Eagle Eye” project, said that after testing the traffic sign recognition database, the designed system “recognition accuracy reached 99.75%, exceeding the current highest accuracy of 99.65%”. A professor of computer vision in 985 colleges and universities, after reading the project brief, told reporters at www.thepaper.cn, "It feels at least the level of an undergraduate graduation thesis. If it is a junior third student, it will be completed independently without any help. It is indeed a little weird, and can be called a prodigy."
Duan Yanyan participated in the National Youth Science and Technology Innovation Contest for three consecutive years from the second day of the second year, and won the third prize of the Youth Science and Technology Innovation Achievement Competition in the 32nd National Youth Science and Technology Innovation Contest The first prize, the first prize of the 34th National Youth Science and Technology Innovation Contest for Youth Science and Technology Innovation Achievements, the winning works are "quantitative analysis and intelligent optimization method of "forkfish in water"", "unmanned eagle eye-based on depth Learning Intelligent Traffic Sign Recognition and APP Implementation", "Using Egg White as a Component-Research and Preparation of Flexible Memristor and Cross Array".
An invention patent application filed earlier than the award-winning work: patent inventor and work tutor with the same name
Duan Mou Yan Gao won the first prize in the 34th National Youth Science and Technology Innovation Contest "Research and Preparation of Egg White as a Component-Flexible Memristor and Cross Array". Tutors were Zhou Guangdong, Huang Shiyou and Li Jiubin .
The project introduction mentioned in the entry mentioned, "Traditional computers are not suitable for processing artificial intelligence tasks that require a large amount of calculation due to the separation of storage and operation, the speed of CPU and memory does not match, the efficiency is low, and the brain is excellent in this respect. Neural network computing and storage are fused and powerful."
"This project uses fresh eggs as raw materials to extract transparent egg white liquid, and proposes a method for modifying egg white proteins using melanin, glycerin, and deionized water. A highly flexible memristor is prepared...with information storage capability."
"Using the egg white flexible memristor, combined with the cross array structure, a foldable memristor cross array is designed, which can be folded and folded, and has a novel structure, which can provide hardware support for the underlying structure of the brain-like computer that realizes the fusion of storage and computing. ."
However, Peng Mei News (www.thepaper.cn) found a patent on the official website of the State Intellectual Property Office for the application of a photoelectric dual-control flexible egg white memristor in the integration of data storage and calculation. Resistors store and process data.
This patent application (patent right) is made by Southwest University, and one of the inventors is Zhou Guangdong, who has the same name as the tutor of the first prize work.
The patent specification mentions that the invention belongs to the field of computer technology, and proposes that in the prior art, in order to deal with the storage and operation requirements of big data, the idea of parallel storage and calculation is proposed. In short, a device that can replace both the CPU and the memory is required, and the binary calculation can be performed directly. The purpose of the invention is to provide an application of photoelectric dual-control flexible egg white memristor in the integration of data storage and calculation.
The technical solution of the invention is that each certain amount of 5,6-dihydroxyindole, 3 polyglycerol, egg white solution and deionized water are mixed in sequence. The egg white memristor finally prepared can not only store data, but also process data.
It is worth noting that the 5,6-dihydroxyindole mentioned in the plan is an important intermediate for melanin, which is the raw material for melanin production; and the polymerization reaction of glycerol molecules can form trimerization, tetramerization, and more Polyglycerol.
In other words, the raw materials used in this patent are highly coincident with the melanin, glycerin, deionized water and egg white protein mentioned in the first prize winning project.
The patent was applied on January 23, 2019, and the 34th National Youth Science and Technology Innovation Contest was filed from March 20 to April 20, 2019. The patent application date was earlier than the contest filing date.
According to the official website of the State Intellectual Property Office, the inventors of the above "Application of a photoelectric dual-control flexible egg white memristor in the integration of data storage and calculation" are Zhou Guangdong, Duan Shukai, Wang Lidan and Hu Xiaofang, and the applicant's address is Tiansheng, Beibei District, Chongqing City. School of Artificial Intelligence, Southwest University, No. 2 Road.
According to the report of the window of Chinese universities in November 2018 named "Three Southwest University Postdoctors Selected for the "Chongqing Postdoctoral Innovative Talent Support Plan for 2018"", Southwest University Postdoctoral Zhou Guangdong entered the 2018 "Chongqing Postdoctoral Innovative Talent Support Plan" List of selected personnel. Zhou Guangdong is engaged in research work in the postdoctoral mathematics mobile station of the school. The cooperative tutor is Professor Duan Shukai, who mainly studies the preparation of memristors, the principle of memristors, and the transitional metal oxides and natural organic materials.
According to the official website of the School of Materials and Energy of Southwest University in May this year, post-doctor Zhou Guangdong participated in the research group on the mapping of memristor evolution maps by Song Qunliang, a doctoral supervisor of the Department of Materials and Energy of Southwest University, and related results were published.
Duan Yanyan is operating experiments at the School of Materials and Energy, Southwest University.
Consistent with the college where Zhou Guangdong participated in the research group, Duan Mouyan's "Operation of Egg White as a Component-Research and Preparation of Flexible Memristor and Cross Array" experimental operation is also in the School of Materials and Energy, Southwest University.
The second inventor of the patent, Duan Shukai, according to the official website of the School of Artificial Intelligence of Southwest University, is currently a professor, doctoral tutor and associate dean of the School of Artificial Intelligence of Southwest University, acting director of the National Engineering Laboratory of Intelligent Transmission and Control Technology, Intelligent Transmission and Control Technology Director of Chongqing Technology Innovation Center, director of Chongqing Key Laboratory of Brain-like Computing and Intelligent Control, deputy director of Brain Cognition Branch Center of Chongqing Brain Science Collaborative Innovation Center. His main research fields include artificial intelligence, brain-like chips, industrial robots, intelligent control, and bionic information materials.
In addition to the patent for "The Application of a Photoelectric Dual-Control Flexible Egg White Memristor in the Integration of Data Storage and Computing", the official website of Southwest University shows that Duan Shukai has collaborated with another inventor of the above patent Wang Lidan on at least 11 national invention patents . Duan Mouyan won the third prize of the 32nd National Youth Science and Technology Innovation Contest in the second year of the "Quantitative Analysis and Intelligent Optimization Method of the "Forkfish in Water" problem. The tutor was named Wang Lidan.
Surging News (www.thepaper.cn) found a professor named Wang Lidan on the official website of the School of Electronic Information Engineering of Southwest University, whose main research directions include artificial neural networks, nonlinear systems and circuit design, memristors and memristors, Bioelectronic circuits and neuromorphic systems.
It is close to Duan Shukai's academic exchanges. The official website shows that the two have cooperated on 22 national invention patents and utility model patents, and only 35 Chinese academic papers have collaborated.
Discussions on the close relationship between Duan Shukai, Duan Mouyan, and Wang Lidan recently sparked discussions on social networks. On July 17, when Peng Mei News (www.thepaper.cn) logged in to the official website of the School of Artificial Intelligence of Southwest University again, it was found that Duan Shukai's "Introduction to Invention Patent" and other content on his profile page have been deleted, including 11 items co-operated with Wang Lidan National invention patent information.
On the personal homepage of Duan Shukai, the official website of the School of Artificial Intelligence of Southwest University, the screenshot of the national invention patent in cooperation with Wang Lidan is 1:13 am on July 16.
The only information currently on Duan Shukai's personal homepage, screenshot time is 12:34 on July 17.
Multi-channel spatial transformation convolutional neural network, recognition accuracy 99.75%? Insiders: Can send top papers on computer vision
In addition to the second day, he won the third prize of the 32nd National Youth Science and Technology Innovation Contest with "Quantitative Analysis and Intelligent Optimization Method of "Forkfish in Water", and Gao Yi used "Use Egg White as a Component-Flexible Memristor and Cross Array" "Research and Preparation" won the first prize of the 34th National Youth Science and Technology Innovation Contest. Duan Mouyan also won the third prize of the 33rd National Youth Science and Technology Innovation Contest in the third year. His work is "Unmanned Eagle Eye-based Deep learning intelligent traffic sign recognition and APP implementation".
Introduction to the project "Unmanned Eagle Eyes-Intelligent Traffic Sign Recognition and APP Implementation Based on Deep Learning".
The project introduction of this work mentions that driverless cars are popular all over the world. "At present, intelligent traffic sign recognition mainly uses deep convolutional neural networks, but it is difficult to meet the speed and accuracy of practical applications because it is easily affected by the diversity of image space. Requirements. In response to this problem, this paper combines a convolutional neural network and a spatial transformation network, and proposes a multi-path spatial transformation convolutional neural network based on deep learning."
"Multi-channel training of input images of different sizes enhances the generalization ability of the network and can well adapt to the spatial diversity of the image. By testing with the German traffic sign recognition database GTSRB, the accuracy of the system design we designed achieves 99.75%, exceeding the current highest accuracy of 99.67%. At the same time, porting the proposed algorithm to the Android mobile phone mobile terminal APP can help pedestrians and vehicles travel safely."
Based on the introduction of the National Youth Science and Technology Innovation Competition official website to the project, an engineer from an autonomous driving algorithm company told Peng Mei News (www.thepaper.cn) that the "system identification accuracy reached 99.75%" mentioned in the entries "If you really exceed other current scholars and achieve 99.75% accuracy, you can publish top-level papers in computer vision, so I feel more hydrated."
According to local media reports, Duan Yanyan was selected as one of the top ten local figures for being a good boy with the dream of a powerful country in science and technology. The reasons for the award include "the personal scientific and technological innovation results won the gold medal in the 34th National Youth Science and Technology Innovation Competition." "Focus on the traffic sign recognition technology in unmanned driving technology, the designed algorithm recognition accuracy reaches 99.75%, exceeding the current highest accuracy of 99.65%."
Duan Yanyan declared three scientific and technological innovation achievements in three years, and each of them won the national award. Among them, "Unmanned Eagle Eyes-Intelligent Traffic Sign Recognition and APP Implementation Based on Deep Learning" and "Using Egg White as a Component-Research and Preparation of Flexible Memristors and Cross Arrays" are national youth science and technology innovation contests Computer science competition.
Although they are all in the field of computer science, a professor of computer vision in a 985 university told Peng Mei News (www.thepaper.cn) that "these two studies are not very relevant." "This project (referring to the unmanned eagle eye project) feels at least the level of an undergraduate graduation thesis. If it is a junior third student who completed it independently without any help, it is indeed a bit incredible and can be called a prodigy."
But he also mentioned that if it is completed under guidance, it is not entirely impossible. "Because it is not clear about the basis of the relevant aspects of this student, how much time is invested, and how to guide, it is not very easy to say."
"It is not difficult to get started with deep learning. With an open source framework, many undergraduates can get started in 2 weeks, and junior high school students can't say that they will not learn, depending on how much time they spend and how well they master." The professor said, "Of course It is not easy for an undergraduate to make innovative things using deep learning, and this project will also be ported to Andriod to make an app, which will require related programming and the workload is not small."
Another person from an artificial intelligence startup told Peng Mei News (www.thepaper.cn) that there are two kinds of people in the field of artificial intelligence. One person can propose algorithms and improve performance through algorithms. "This kind of person needs to have deeper knowledge accumulation, because deep learning is more clearly a system engineering, which has high requirements for mathematics, computers, and English. Even if junior high school students can know the theory, but systematically implement a project. Not optimistic, unless it is really a genius." The second kind is "parameter engineer", that is, try to adjust some parameters on the framework of others to get better results. "As long as you know the evaluation standard, set the operating environment, and then run the program Just do it."
Introduction of the project "Using Egg White as a Component-Research and Preparation of Flexible Memristor and Cross Array".
In addition, an industry person working in a domestic chip R&D company gave a brief introduction to Duan Yanyan's first prize work "Research and Preparation of Egg White as a Component-Flexible Memristor and Cross Array" (www.thepaper .cn) said, "I think this technology is not innovation, but subversion." In the project introduction, "Each sentence in the back is a worldwide problem in biology, computers, chips, and artificial intelligence."