The second main contribution: An original model of "visual perception-eye movement control neural network" automatic target search and recognition system is proposed. This model is different from the traditional traversal target search and recognition model. Advanced perceptual cognitive neurons sparsely encode the contextual spatial relationship of image targets, and then infer and identify the location and category of targets.
The above research contents were published in nearly 30 international and domestic journals and conference papers respectively. Among them, 4 were retrieved by SCI, 5 by ISPEC and 4 by EI 1 1. Four articles were cited by SCI for 23 times, nine articles by CSCD for 25 times, and a single article published in the internationally renowned publication Pattern Recognition was cited by SCI 16 times in 2006. Up to 2007, 1 1 articles were cited by Google academically 152 times, among which a single article published in the internationally renowned publication Pattern Recognition was cited 1 1 times. Eight patents have been applied for related technologies, one of which was authorized in 2006. During my master's and doctoral studies, I won seven awards from the state, Shanghai, foreign-funded research institutes, Chinese Academy of Sciences and universities, including Microsoft Scholarship (2000) and Dean's Scholarship of Chinese Academy of Sciences (2005).
Cooperate and exchange with the Graduate School of China Academy of Sciences, Beijing University of Technology, Ohio State University, Michigan State University, Hubei Ezhou Lianhuashan Computer Vision Research Institute, and Beijing Shengkun Yanlun Technology Co., Ltd. Visual Arts and Brain Cognition Laboratory, and preside over or jointly take charge of 5 scientific research projects. The main research contents include: the mechanism of visual neurodynamics (pulse synchronous vibration) and visual neuroinformatics (feature extraction and coding memory). Bottom-up visual perception and eye movement attention mechanism (feature selection and binding, target segmentation and recognition), top-down visual cognition and action control mechanism (image understanding and reconstruction, target search and tracking), visual neural network construction, neural information coding, machine learning and its application in image processing and pattern recognition, such as automatic segmentation, search, detection, tracking, recognition and reconstruction of image targets such as faces, pedestrians, vehicles and people.