News&Events

Focus

​Highly Cited (ESI) Papers Published by BIT Research Teams are Grouped Together in the Field of Computer Science Publication among the Yearly Top 1% of the Best

  According to the data of Essential Science Indicators in Web of Science, the SCI papers published by the research team of Professor Xu Tingfa of the School of Optoelectronics of BIT became highly cited papers in the ESI database. This data shows that by 2020 January/February, according to the citation threshold level, the number of citations received for this paper has been classified in the top 1% in the field of Computer Science, and its academic value has been concerned and affirmed by global researchers in this field. 
      Research paper: Scale-Aware Fast R-CNN for Pedestrian Detection, (First Author: Dr. Li Jianan; Corresponding Author: Professor Xu Tingfa), published in IEEE Transactions on Multimedia, the journal published by the American Institute of Electrical and Electronics Engineers (IEEE), The impact factor in the past 5 years was 5.551.
      The thesis aims at the different feature patterns of pedestrian targets at different scales, and the intra-class differences caused by the input target scale changes will greatly increase the learning difficulties and challenges of the target detection model. The thesis takes the strategy of treating different objectives differently, while sharing the underlying common features to ensure computational efficiency, the exclusive network branch is used to capture the unique high-level features of different scale targets. Independent prediction and scale adaptive fusion are performed to enhance the model's robustness to input scale changes as well. This method is simple and easy to implement, at the same time it can be effectively applied to pedestrian detection in real complex scenes. More importantly, this method provides an effective and practical new idea to improve the robustness of the detection model against input changes, and is widely used in other tasks in the field of computer vision.

  

Figure: Proposed Scale-aware Pedestrian Small Target Detection Framework

  

Introduction of the first author:
      Li Jianan, a 2013th doctoral student at the School of Optoelectronics of BIT, learned after Professor Xu Tingfa. His research directions are optical imaging detection & recognition, image & video processing, computer vision, and multimedia. In 2015, he was sponsored by the National High-level Public Study Abroad Program, going to the National University of Singapore (NUS) for a two-year joint training for doctoral students. The foreign tutors are famous scholars, Yan Shuicheng and Professor Feng Jiashi, in the field of computer vision.
      During his Ph.D. study, he won the Wang Daheng University Student Optical Award, the National Scholarship for Ph.D. Graduate Students, the High-level Doctoral Dissertation Fund of BIT, the Excellent Graduate Degree Thesis of BIT, and many other innovative talent awards. He also won the 2017 ImageNet International Large-scale The champion of the team of visual recognition challenge target positioning task. A total of 16 papers have been published in SCI journals such as IEEE TPAMI, TVCG, TCSVT, TMM, and top international conferences such as CVPR, ICLR, and NeurIPS, including 9 first authors, with a total of more than 1,200 citations. At the same time, he has also served as a reviewer for many international SCI journals such as IEEE TCSVT, TMM and ICCV 2019, CVPR 2020, ECCV2020 and other international top conferences.

  
Corresponding author profile:

  Xu Tingfa, Professor, Ph.D. supervisor, responsible professor of a national first-level key discipline "Optical Engineering", and deputy director of the Key Laboratory of Optoelectronic Imaging Technology and Systems, Ministry of Education. In recent years, he has led its scientific research team to continuously deepen its research around photoelectric imaging detection and recognition, and hyperspectral computing imaging processing. Presided over 30 major scientific research instruments and other special projects of the National Natural Science Foundation of China. He has published more than 120 academic papers in international and domestic journals, among which more than 80 have been included in SCI/EI. He applied for 15 national invention patents as the first inventor, 8 of which have been authorized and publicized. He won a second prize for national defense scientific and technological progress, a second prize for military scientific and technological progress, and a third prize for Beijing science and technology progress award. The innovative research results obtained by the scientific research team have contributed to the construction of the school's first-class disciplines.