Song Chengqun, graduated from the College of Electronics and Information Engineering, Sichuan University, and obtained a doctoral degree. He completed his postdoctoral research at the Chinese Academy of Sciences. He was a visiting scholar at Jeonbuk National University in South Korea. Currently, he holds the position of Associate Professor and Master's Supervisor at the Institute of Advanced Technology, Chinese Academy of Sciences, Shenzhen. He is recognized as a high-level professional talent in Shenzhen and a leading talent in Nanshan District, Shenzhen. He has received funding from the China Postdoctoral Science Foundation. As a key member, he has participated in the National Key Research and Development Program of the Ministry of Science and Technology, the Key Area Research and Development Program of the Guangdong Provincial Department of Science and Technology, joint funds and general projects of the National Natural Science Foundation of China, and has made outstanding contributions in areas such as multi-robot cooperative motion, multi-robot calibration systems, 3D scene reconstruction, and perception. He has led basic research projects and sub-projects of key basic research projects in Shenzhen. He has participated in several collaborative projects commissioned by companies such as China Yangtze Power, China General Nuclear, Tencent, and Mount Tai Sports, and has developed a series of intelligent product prototypes including mobile robots, vision detection systems, and nuclear power plant interaction systems. He has been granted 15 invention patents and has published over 20 high-level papers. His main research areas include computer vision, deep learning, SLAM, and robotics.
Research Areas
computer vision, deep learning, SLAM, robotics
Publications
Papers
[1] Chengqun Song, Maolong Niu, Zhipeng Liu, Jun Cheng, Peng Wang, Hongjian Li, Luoying Hao, Spatial-temporal 3D dependency matching with self-supervised deep learning for monocular visual sensing, Neurocomputing, 481(7): 11-21, 2022.
[2] Chengqun Song, Bo Zeng, Tong Su, Ke Zhang, Jun Cheng, Data association and loop closure in semantic dynamic SLAM using the table retrieval method, Applied Intelligence, 52: 11472-11488, 2022.
[3] Tong Su, Shuo Ye, Chengqun Song*, Jun Cheng. Mask-Vit: An Object Mask Embedding in Vision Transformer for Fine-Grained Visual Classification, 29th IEEE International Conference on Image Processing, ICIP 2022, October 16, 2022 - October 19, 2022, Bordeaux, France: 1626-1630, 2022.
[4] Tong Su, Chengqun Song*, Jun Cheng. Vision Transformer Withinformation Bottleneck Forfine-Grained Visual Classification International Conference on Guidance, Navigation and Control, ICGNC 2022, August 5, 2022 - August 7, 2022, Harbin, China,845 LNEE: 4010-4019, 2023.
[5] Bo Zeng, Chengqun Song*, Cheng Jun, Yuhang Kang, DFPC-SLAM: A Dynamic Feature Point Constraints-Based SLAM Using Stereo Vision for Dynamic Environment, Guidance, Navigation and Control, 3(1): 2350003, 2023.
[6] Chengqun Song, Jun Cheng, Wei Feng, A Crowdsensing-based real time system for finger interactions in intelligent transport system, Wireless Communications and Mobile Computing, 2017: 7385052, 2017.
[7] Chengqun Song, Jun Cheng, A robust projector–camera interactive display system based on finger touch control by fusing finger and its shadow, Journal of the Society for Information Display, 25(9): 568-576, 2017.
[8] Hushan Qin, Jun Cheng, Chengqun Song, Fusheng Hao, Qin Cheng. Structure-Preserving View-Invariant Skeleton Representation for Action Detection, 26th International Conference on Pattern Recognition, ICPR 2022, August 21, 2022 - August 25, 2022, Montreal, QC, Canada, 2022-August: 3190-3196, 2022.
[9] 宋呈群，程俊，曾驳，苏童，张可，张锲石，基于深度扭曲的无监督深度学习视觉SLAM，2022国际制导、导航与控制学术会议（ICGNC2022），2022.
[10] Zhen Liu, Jun Cheng, Libo Liu, Ziliang Ren, Qieshi Zhang, Chengqun Song, Dual-stream cross-modality fusion transformer for RGB-D action recognition, Knowledge-Based Systems, 255(14):109741, 2022.
[11] Zhen Liu, Qin Cheng, Chengqun Song, Jun Cheng. Cross-Scale Cascade Transforer for Multimodal Human Action Recognition. Pattern Recognition Letters, 168: 17-23, 2023.
[12] Fusheng Hao, Fuxiang Wu, Fengxiang He, Qieshi Zhang, Chengqun Song, Jun Cheng, Semantic-Aware Feature Aggregation for Few-Shot Image Classification, Neural Processing Letters, DOI: 10.1007/s11063-023-11150-2, 2023.
[13] 李雪松，张锲石，宋呈群，康宇航，程俊，自动驾驶场景下的轨迹预测综述，计算机工程，49(5): 1-11, 2023.
Patents
[1] Chengqun Song, Jun Cheng. Method and module for constructing 3D semantic maps, and robot interaction method and system. December 8, 2020. China. Patent number: CN202011442377.4.
[2] Chengqun Song, Jun Cheng. Method and device for simultaneous localization and mapping based on appearance retrieval. August 7, 2020. China. Patent number: CN202010787859.7.
[3] Chengqun Song, Jun Cheng, Peng Wang. Training method, device, and terminal equipment for neural network for scene understanding. June 28, 2019. China. Patent number: CN201910573347.8.
[4] Bo Zeng, Chengqun Song, Jun Cheng. Visual SLAM method, device, terminal equipment, and storage medium. November 9, 2021. China. Patent number: CN202110805748.9.
[5] Jun Cheng, Chengqun Song, Bo Zeng, Fuxiang Wu. Robot control method and robot. August 13, 2021. China. Patent number: CN202110428814.5.
[6] Jun Cheng, Ying Hu, Chengqun Song, Yuanyuan Yang. Robot calibration method and terminal. October 28, 2019. China. Patent number: CN201911028907.8.
[7] Tong Su, Chengqun Song, Jun Cheng. Image classification method based on convolutional neural network-transformer hybrid architecture. April 15, 2022. China. Patent number: CN202111479058.5.
[8] Ke Zhang, Chengqun Song, Jun Cheng, Bo Zeng, Hushan Qin, Zhen Liu. Inspection path processing method, device, equipment, and readable storage medium. July 7, 2022. China. Patent number: CN202210794421.0.
[9] Chengqun Song, Jun Cheng, Fuxiang Wu, Haiguang Guo, Xiangyang Gao. Guidance method, device, and readable storage medium. June 24, 2022. China. Patent number: CN202210722609.4.
[10] Zhen Liu, Jun Cheng, Ziliang Ren, Chengqun Song, Qieshi Zhang. Multimodal information fusion recognition method and system based on attention mechanism. April 12, 2022. China. Patent number: CN202111557072.2.
[11] Yichao Zhou, Jun Cheng, Peng Wang, Chengqun Song, Bo Cheng, Yu Luan, Yiqian Wu, Xuegang Zhang, Zhenhua Luan. Gesture control method, device, equipment, and storage medium for nuclear power plant control system. April 2, 2021. China. Patent number: CN202011260724.1.
[12] Yu Luan, Chengqun Song, Jun Cheng, Bo Cheng, Yichao Zhou, Yiqian Wu, Xuegang Zhang, Zhiyao Liu, Zhenglong Yu, Zhenhua Luan. Method, device, equipment, and medium for obtaining depth images based on deep learning. December 4, 2019. China. Patent number: CN201911225113.0.
[13] Bo Cheng, Jun Cheng, Chengqun Song, Peng Wang, Yichao Zhou, Yu Luan, Yiqian Wu, Xuegang Zhang, Zhenglong Yu, Zhiyao Liu, Jianbo Zhang, Xiaomei Xu, Zhenhua Luan. Multichannel Fusion Human-Machine Interaction Method and System for Nuclear Power Plant Control Room, August 8, 2019, China, CN201910729952.X
[14] Chengqun Song, Jun Cheng. Panoramic 3D video stitching system and method. December 13, 2016. China. Patent number: CN201611146790.X.
[15] Jun Cheng, Fuxiang Wu, Chengqun Song. Text-based image editing method and electronic device. February 25, 2022. China. Patent number: CN202111188395.9.
Students
现指导学生
刘桢 硕士研究生 085400-电子信息
秦虎山 硕士研究生 085400-电子信息
曾驳 硕士研究生 085400-电子信息