研究領域
1. 計算機視覺和機器學習
--Object Recognition andSegmentation
--Deep Learning and Image Stylization
--手寫體識別, 車牌識別等
2. 醫學影像分析
--Deep Learning and Medical Image Segmentation
3. 金融大數據分析
--時序分析, 統計分析, 模式分析等
科研項目
From Image Parsing to Painterly Rendering (TOG)Painterly Animation with Video Content Extraction (NPAR, TVCG)Computer Generated Embroidery
代表論著
Major Journals
Liang Lin, Ping Luo, Xiaowu Chen, andKun Zeng, "Representing and Recognizing Objects with Massive Local Image Patches", Pattern Recognition, accepted, 2011.
Kun Zeng, Mingtian Zhao, Caiming Xiong and Song-Chun Zhu, "From Image Parsing to Painterly Rendering", ACM Transactions on Graphics. vol. 29 (1), Article 2, 2010 (Cover Paper, Presented in Siggraph2010)
Kun Zeng, Liang Lin, Huaiyu Wu, Chunhong Pan, Qing Yang, "Automatic Natural Image Sketching with Low-, Mid-level Cues", Journal of Information and Computational Science, 5(4): 1647-1655, 2008
Liang Lin,Kun Zeng, Yongtian Wang, and Wenze Hu,"3D Structure Inference by Integrating Segmentation and Recnstruction from a Single Image", IET Computer Vision, 2(1): 15-22, 2008.
Yu Wu, Chunhong Pan,Kun Zeng, Qing Yang, Sngde MA. "An Efficient Skeleton-Free Pose Retargetting Method for Triangular Meshes". Chinese Journal of Electronics (English Version), 4, pp. 659-664, 2008.
Liang Lin, Yongtian Wang, Caiming Xiong , Xiaoming Hu,Kun Zeng, "Marker-less Registration Based on Template Tracking for Augmented Reality", Multimedia Tools and Applications, vol. 41(2), pp. 235-252, 2009.
Liang Lin,Kun Zeng, "Marker-less Registration Based on Template Tracking for Augmented Reality", Multimedia Tools and Applications, vol. 41(2), pp. 235-252, 2009.