Tue AM1.L4: Interpretable and Explainable Deep Learning
Tue, 5 Sep, 11:00 - 13:00 Finland Time (UTC +3)
Location: Nautica
Session Type: Lecture
Session Chair: Nikos Deligiannis, Vrije Universiteit Brussel
Track: SiG-DML - Signal and Data Analytics for Machine Learning
Tue, 5 Sep, 11:00 - 11:20 Finland Time (UTC +3)

Tue AM1.L4.1: STOCHASTIC UNROLLED PROXIMAL POINT ALGORITHM FOR LINEAR IMAGE INVERSE PROBLEMS
Brandon Le Bon, INRIA Rennes - Bretagne Atlantique, France; Mikaël Le Pendu, INTERDIGITAL, France; Christine Guillemot, INRIA Rennes - Bretagne Atlantique, France
Tue, 5 Sep, 11:20 - 11:40 Finland Time (UTC +3)

Tue AM1.L4.2: PRIOR FOR MULTI-TASK INVERSE PROBLEMS IN IMAGE RECONSTRUCTION USING DEEP EQUILIBRIUM MODELS
Samuel Willingham, Inria, France; Mårten Sjöström, Mid Sweden University, Sweden; Christine Guillemot, Inria, France
Tue, 5 Sep, 11:40 - 12:00 Finland Time (UTC +3)

Tue AM1.L4.3: Variance Predictions in VAMP/UAMP with Right Rotationally Invariant Measurement Matrices for niid Generalized Linear Models
Zilu Zhao, Dirk Slock, EURECOM, France
Tue, 5 Sep, 12:00 - 12:20 Finland Time (UTC +3)

Tue AM1.L4.4: GRAPH-TIME TREND FILTERING AND UNROLLING NETWORK
Mohammad Sabbaqi, Elvin Isufi, Delft University of Technology, Netherlands
Tue, 5 Sep, 12:20 - 12:40 Finland Time (UTC +3)

Tue AM1.L4.5: Quantitative Evaluation of Video Explainability Methods via Anomaly Localization
Xinyue Zhang, Boris Joukovsky, Nikos Deligiannis, Vrije Universiteit Brussel, Belgium
Tue, 5 Sep, 12:40 - 13:00 Finland Time (UTC +3)

Tue AM1.L4.6: HARNESSING THE POWER OF EXPLANATIONS FOR INCREMENTAL TRAINING: A LIME-BASED APPROACH
Arnab Neelim Mazumder, University of Maryland Baltimore County, United States; Niall Lyons, Ashutosh Pandey, Avik Santra, Infineon Technologies, United States; Tinoosh Mohsenin, University of Maryland Baltimore County, United States