Ngai-Man (Man) Cheung
Associate Professor
Singapore University of Technology and Design (SUTD)

E-mail: ngaiman_cheung(at)sutd.edu.sg

Tel: +65 6499 4542

Building 1, Level 5, 1.502-17 
8 Somapah Road, Singapore 487372
Research Interest: Image and Signal Processing/Analysis, Computer Vision, Machine Learning, AI


  • 09/2019: Self-supervised GAN paper in NeurIPS-19 (NIPS-19) 
    • Trung Tran, Hung Tran, Ngoc Nguyen, Linxiao Yang, Ngai-Man Cheung, "Self-supervised GAN: Analysis and Improvement with Multi-class Minimax Game," in Proc. NeurIPS-19 (NIPS-19) (Total 6743 submissions. 21.1% acceptance rate) [PDF]
  • 09/2019: Deep clustering paper in ICCV-19
    • Linxiao Yang, Ngai-Man Cheung, Jiaying Li, and Jun Fang, "Deep Clustering by Gaussian Mixture Variational Autoencoders with Graph Embedding," in Proc. ICCV-19 [PDF]
  • 06/2019: CVPR-19 Best Paper Finalist (Total 5160 submissions)
    • H Le, TT Do, T Hoang, NM Cheung, "SDRSAC: Semidefinite-Based Randomized Approach for Robust Point Cloud Registration without Correspondences," in Proc. IEEE Conference on Computer Vision and Pattern Recognition (CVPR-19) (Oral) [PDF]
  • 04/2019: Recent papers on 3D computer vision
    • H Le, TT Do, T Hoang, NM Cheung"SDRSAC: Semidefinite-Based Randomized Approach for Robust Point Cloud Registration without Correspondences," in Proc. IEEE Conference on Computer Vision and Pattern Recognition (CVPR-19) (Oral) [PDF]
    • NT Tran, DK Le Tan, AD Doan, TT Do, TA Bui, M Tan, NM Cheung, "On-Device Scalable Image-Based Localization via Prioritized Cascade Search and Fast One-Many RANSAC." IEEE Transactions on Image Processing. [PDF]
  • 11/2018: Improve training of GAN
    • Ngoc-Trung Tran, Tuan-Anh Bui, Ngai-Man Cheung, "Improving GAN with neighbors embedding and gradient matching," in Proc. The Thirty-Third AAAI Conference on Artificial Intelligence (AAAI-2019) [PDF] (Total 7700 submissions. 16.2% acceptance rate)
  • 9/2018: Anomaly detection with GAN
    • SK Lim, Y Loo, NT Tran, Ngai-Man Cheung, G Roig, Y Elovici, "DOPING: Generative Data Augmentation for Unsupervised Anomaly Detection with GAN," in Proc. IEEE International Conference on Data Mining 2018 (ICDM-2018) [PDF]
  • 9/2018: Activity recognition in video with self attention
    • S Song, Ngai-Man Cheung, V Chandrasekhar, B Mandal, "Deep Adaptive Temporal Pooling for Activity Recognition," in Proc. ACM Multimedia 2018 (Full Research Paper) [PDF]
  • 7/2018: An improved GAN
    • Ngoc-Trung Tran, Tuan-Anh Bui, Ngai-Man Cheung"Dist-GAN: An Improved GAN using Distance Constraints," in Proc. European Conference on Computer Vision (ECCV 2018) [PDF] [Supplementary]
  • 6/2018: Paper on multi-level similarity
    • Yiluan Guo, Ngai-Man Cheung“Efficient and Deep Person Re-Identification using Multi-Level Similarity,” in Proc. IEEE Conference on Computer Vision and Pattern Recognition (CVPR-18). [PDF]
  • 2018: Paper on deep neural network compression
    • Yiren Zhou, Seyed-Mohsen Moosavi-Dezfooli, Ngai-Man Cheung, Pascal Frossard, “Adaptive Quantization for Deep Neural Network,” in Proc. Thirty-Second AAAI Conference on Artificial Intelligence (AAAI-18). [PDF]
  • 2017: Paper on sparse Graph Fourier Transform (GFT) for network traffic anomalies detection
    • Manas Khatua, Seyed Hamid Safavi, Ngai-Man Cheung, "Sparse Laplacian Component Analysis for Internet Traffic Anomalies Detection." IEEE Transactions on Signal and Information Processing over Networks, 2017. [PDF]


I received my Ph.D. degree in Electrical Engineering from 
University of Southern California (USC), Los Angeles, CA, in 2008. My Ph.D. research focused on image and video coding, and the work was supported in part by NASA-JPL.

From 2009-2011, I was a postdoctoral researcher with the Image, Video and Multimedia Systems group at Stanford University, Stanford, CA. From 2012-2018, I was an Assistant Professor with Singapore University of Technology and Design (SUTD).

I have also held research positions with Texas Instruments Research Center Japan, Nokia Research Center, IBM T. J. Watson Research Center, HP Labs Japan, Hong Kong University of Science and Technology (HKUST), and Mitsubishi Electric Research Labs (MERL). My research has resulted in 11 U.S. patents granted with several pending, 2 technology licensing and 1 start-ups.