Program (Full Day – October 11, 2021) – Eastern Daylight Time (EDT)


8:00-09:50 Session 1:  PCA and Low Rank Recovery (Chair: Prof. Soon Ki Jung)


8:00-8:05 Opening (5 min) (Prof. Soon Ki Jung)

8:05-9:00 Unlabeled Principal Component Analysis (Invited Talk) (45 min)

Manolis Tsakiris, School of Information Science and Technology, ShanghaiTech University, China)

9:00-9:25 Relaxations for Non-Separable Cardinality/Rank Penalties (20 min)

C. Olsson (Lund University, Sweden), Daniele Gerosa (Lund University), M. Carlsson (Lund University)

9:25-9:50 Double-Weighted Low-Rank Matrix Recovery Based on Rank Estimation (20 min)

Zhenqin Xu (WUST), H. Xing (WUST), S. Fang (WUST), S. Wu (School of Machinery and Automation, WUST), S. Xie (Institute for Infocomm Research, Singapore)


11:00-12:30 Session 2Deep Subspace Learning and Clustering (Chair: Prof. Thierry Bouwmans)


11:00-11:55 A Critical View of Self-Expressive Deep Subspace Clustering (Invited Talk) (45 min)

Ben Haeffele, Johns Hopkins Mathematical Institute for Data Science, USA

11:55-12:25 Convolutional Auto-Encoder with Tensor-Train Factorization (20 min)

Manish Sharma (RIT), P. Markopoulos (RIT), E. Saber (RIT), M. Salman Asif (University of California, Riverside), A. Prater-Bennette (Air Force Research Laboratory)

12:25-12:30 Discussion (5 min)

12:30-13:30 Break 2 (Lunch) – 1h


13:30-15:30 Session 3: Robust PCA and its Applications in Computer Vision (Chair: Prof. Mohamed Shehata)


13:30-14:25 Deep Unfolded Robust PCA: Method and Applications (45 min)

Nikos Deligiannis, Vrije Universiteit Brussel, IMEC, Belgium, and  Yonina C. Eldar, Weizmann Institute of Science, Israël

14:25-14:50 Fast Robust Tensor Principal Component Analysis via Fiber CUR Decomposition (20 min)

H. Cai (University of California, Los Angeles), Zehan Chao (University of California, Los Angeles), L. Huang (University of California, Los Angeles), D. Needell (UCLA)

14:50-15:15 TransBlast: Self-Supervised Learning Using Augmented Subspace with Transformer for Background/Foreground Separation (20 min)

Islam Osman (University of British Columbia), M. Abdelpakey (University of British Columbia), M. Shehata (University of British Columbia)

15:15-15:50 – Open Discussion

15:50-16:30 – Break 3 (Coffee) – 40 min


16:30-18:00 Session 4 : Background/Foreground Separation and Anomaly Detection (Chair: Prof. Panos Markopoulos)

16:30-17:00 Background/Foreground Separation: Guided Attention based Adversarial Modeling (GAAM) versus Robust Subspace Learning Methods (20 min)

Maryam Sultana (Kyungpook National University), A. Mahmood (Information Technology University), T. Bouwmans (Univ. La Rochelle), M. Haris Khan (Muhammad Bin Zayed University of Artificial Intelligence), S . Jung (Kyungpook National University)

17:00-17:30 Graph CNN for Moving Object Detection in Complex Environments from Unseen Videos (20 min)

Jhony Giraldo (Univ. La Rochelle), S. Javed (Khalifa University of Science and Technology), N. Werghi (Khalifa University of Science and Technology), T. Bouwmans (Univ. La Rochelle)

17h30-18:00  Synthetic Temporal Anomaly Guided End-to-End Video Anomaly Detection (20 min)

Marcella Astrid (University of Science and Technology), M. Zaigham Zaheer (University of Science and Technology), S. Lee (Electronics and Telecommunication Research Institute)