主讲课程:
程序设计语言原理（本科生课程）
机器学习（本科生课程）
机器学习（硕博生课程）
计算机视觉（博士生课程）
在研项目:
低空航拍下基于隐式姿态模型的平躺人体检测方法研究,青年科学基金项目,2013-2015
代表性论文:
苏松志*，李绍滋，蔡国榕．《行人检测：原理与应用》，厦大出版社-南强丛书，2016-3.
song-zhi su, sin-sian wu, shu-yuan chen, der-jhy duh, and shao-zi li, multi-view fall detection based on spatio-temporal interest points, multimedia tools and applications, 2015, vol.74, no. 14, 15 july. doi 10.1007/s11042-015-2766-3.
song-zhi su, zhi-hui liu, su-ping xu, shao-zi li, rongrong ji. depth image feature learning based on sparse auto-encoder for human detection, signal processing, 2015, 112:43-52. (sci, jcr 3)
song-zhi su, zong-yu lan, shao-zi li, shu-yuan chen, category-specific scene categorization, journal of the chinese institute of engineers, 2015, 38(1):128-137. (sci, jcr 4)
song-zhi su*, shu-yuan chen. analysis of feature fusion based on hik svm and its application for pedestrian detection. abstract and applied analysis. 2013. （sci）volume 2013, page: 1-11, http://dx.doi.org/10.1155/2013/436062.
li-chuan geng, pierre-mar jodoin , song-zhi su*, shao-zi li, cbdf: compressed binary discriminative feature, neurocomputing, 2015, doi:10.1016/j.neucom.2015.07.120. (sci, jcr 3区)
guo-rong cai, pierre-marc jodoin, shao-zi li, yun-dong wu, song-zhi su*, zhen-kun huang. perspective-sift: an efficient tool for low-altitude remote sensing image registration. signal processing, 2013. (sci)
dao-xun xia, song-zhi su*, shao-zi li, pierre-mar jodoin. lying-pose detection with training dataset expansion. icip, 2014.
de-dong yuan, jie dong, song-zhi su*, shao-zi li, rong-rong ji. pursuing detector efficiency for simple scene pedestrian detection, the 20th anniversary international conference on multimedia modeling dublin, ireland, 2014.
li-chuang geng, song-zhi su, dong-lin cao, shao-zi li. perspective-invariant image matching framework with binary feature descriptor and apso. international journal of pattern recognition and artificial intelligence, 2014, 28(8): 1-18. doi: 10.1142/
si chen, shaozi li, song-zhi su, donglin cao, rongrong ji. online semi-supervised compressive coding for robust visual tracking, journal of visual communication and image representation, 2014,25:793-804.
si chen, shaozi li, songzhi su, qi tian, rongrong ji. online mil tracking with instance-level semi-supervised learning, neruocomputing, 2014,139:272-288.
hongbo zhang, shang-an li, shu-yuan chen, song-zhi su, der-jyh duh, and shaozi li, “adaptive photograph retrieval method,” multimedia tools and applications, vol. 70, no. 3, pp. 2189-2209, june, 2014. (sci, ei)
bing shuai, cheng yun, shao-zi li, song-zhi su. a hierarchical clustering based non-maximum suppression method in pedestrian detection. intelligent science and intelligent data engineering. springer berlin heidelberg, 2012. 201-209.
hong-bo zhang, song-zhi su, shao-zi li, duan-sheng chen, bineng zhong, rongrong ji. seeing actions through scene context. vcip, 2013, 11.17-20.
苏松志*, 李绍滋, 陈淑媛, 蔡国榕, 吴云东. 行人检测技术综述[j]. 电子学报. 2012, 40(4): 814-820. (ei)
蔡国榕，*李绍滋，吴云东，苏松志，陈水利. 一种透视不变的图像匹配算法，自动化学报，2013, 39(7): 1053-1061.
zhang h, *li s, chen s, su s, lin x, cao d, 2013. locating and recognizing multiple human actions by searching for maximum-score subsequences, signal, image and video processing, doi: 10.1007/s11760-013-0501-y.
*su s, chen s, 2013. analysis of feature fusion based on hik svm and its application for pedestrian detection, abstract and applied analysis, 2013 (2013), article id: 436062.
zhang h, *li s, su s, chen s, 2013. selecting effective and discriminative spatio-temporal interest points for recognizing human action, ieice transactions on information and systems, e96-d(8): 1783-1792.
cai g, pierre-marc j, li s, wu y, su s, huang z, 2013. perspective-sift: an efficient tool for low-altitude remote sensing image registration, signal processing, 93(11): 3088-3110.