Program


8:00-09:50 Session 1:  Subspace Learning and Applications


8:00-8:05 Opening (5 min)

8:05-8:50 Invited Talk 

L1-norm Analysis of Matrices and Tensors: Theory, Algorithms, and Some Applications (45 min)

Panos P. Markopoulos (Rochester Institute of Technology, USA)

8:50-9:10 Learning Disentangled Representations via Independent Subspaces (20 min)

Maren Awiszus (Leibniz Universität Hannover), Hanno Ackermann (Leibniz University Hannover), Bodo Rosenhahn (Leibniz University Hannover)

9:10-9:30 Uncalibrated Non-Rigid Factorisation by Independent Subspace Analysis (20 min)

Sami Brandt (IT University of Copenhagen), Hanno Ackermann (Leibniz University Hannover), Stella Graßhof (Leibniz Universität Hannover)

9:30-09:50 Robust Discrimination and Generation of Faces using Compact, Disentangled Embeddings (20 min)

Björn Browatzki (Korea University) Christian Wallraven (Korea University)

10:00-11:00 Break 1 (Coffee) – 1h


11:00-12:30 Session 2:  Tensor based  Subspace Learning Approaches


11:00-11:30 Invited Talk

TBD (30 min)

Jean Kossaifi (Samsung AI Cambridge, UK)

11:30-11:50 Tensor Subspace Learning and Classification: Tensor Local Discriminant Embedding for Hyperspectral Image (20 min)

Lei He (Beijing University of Chemical Technology), Hongwei Yang (Beijing University of Chemical Technology), Lina Zhao (Beijing University of Chemical Technology) 

11:50-12:10 Tensor Train Decomposition for Efficient Memory Saving in Perceptual Feature-maps (20 min)

Taehyeon Kim (Yonsei University), Jieun lee (Yonsei University), Yoonsik Choe (Yonsei University)

12:10-12:30 Low-rank Tensor Tracking  (20 min)

Sajid Javed (Khalifa University of Science and Technology), Jorge Manuel Miranda Dias (Khalifa University of Science and Technology), Naoufel Werghi (Khalifa University of Science and Technology)

12:30-13:30 Break 2 (Lunch) – 1h


13:30-15:30 Session 3: Bridging Subspace Learning/Clustering and Deep Learning


13:30-14:10 Invited Talk

Bridging Robust PCA and Variational Autoencoder Models (40 min)

David Wipf (Microsoft Research, China)

14:10-14:30 Structuring Autoencoders (20 min)

Marco Rudolph (Leibniz University Hannover), Bastian Wandt (Leibniz University Hannover), Bodo Rosenhahn (Leibniz University Hannover)

14:30-14:50 Structure-Constrained Feature Extraction by Autoencoders for Subspace Clustering (20 min)

Kewei Tang (Liaoning Normal University), Kaiqiang Xu  (Dalian University of Technology), Zhixun Su (Dalian University of Technology), Wei Jiang (Liaoning Normal University),  Xiaonan Luo (Guilin University of Electronic Technology), Xiyan Sun (Guilin University of Electronic Technology)

14:50-15:10 Deep Closed-Form Subspace Clustering (20 min)

Junghoon Seo (Satrec Initiative), Jamyoung Koo (SI Analytics), Taegyun Jeon (SI Analytics)

15:10-15:30 Subspace Clustering with Missing Entries byNon-convex Minimization (20 min)

Connor Lane (John Hopkins University), Ron Boger (John Hopkins University), Chong You (John Hopkins University), Manolis Tsakiris (ShanghaiTech University), Benjamin D. Haeffele (John Hopkins University), René Vidal (John Hopkins University)

15:30-15:50 Adaptive online k-subspaces with cooperativere-initialization

Connor Lane (John Hopkins University), Benjamin D. Haeffele (John Hopkins University), René Vidal (John Hopkins University)

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


16:30-18:00 Session 4 : Application in Computer Vision: Background/Foreground Separation


16:30-17:00 Invited Talk

TBD (30 min)

Sajid Javed (Khalifa University of Science and Technology)

17:00-17:20 Panoramic Video Separation with Online Grassmannian Robust Subspace Estimation (20 min)

Kyle Gilman (University of Michigan), Laura Balzano (University of Michigan)

17:20-17:40 Topological Labelling of Scene using Background/Foreground Separation and Epipolar Geometry (20 min)

Atsushi Imiya (IMIT Chiba University, Japan), Hiroki Hiraoka (Chiba University)

17h40-18:00 Complete Moving Object Detection in the Context of Robust Subspace Learning (20 min)

Maryam Sultana (Kyungpook National University), Arif Mahmood (Information Technology University), Thierry Bouwmans (La Rochelle Université), Soon Ki Jung (Kyungpook National University)