Adapting to Change: Reliable
Learning Across Domains
ECML-PKDD 2023 Workshop
Workshop Program
The plan of the workshop is to have a combination of invited talks from leading researchers, that could address the topic of the workshop from different perspectives, together with a call for papers to be presented within a poster session, in order to create an interactive environment for the attendees. In our experience, this is the most effective format for discussing new ideas with participants interested in the papers and making meaningful and lasting connections. Invited speakers will also be encouraged to take part in the poster session and provide their valuable feedback to authors. We also plan to dedicate time slots to questions after each invited talk to foster interaction between the speakers and the attendees.
Important Information:
The workshop will take place from 2.30PM on Monday, September 18th, 2023.
If you plan to present a poster, please be at the venue at 2PM to set it up in advance
Venue: Room 1T @ Politecnico di Torino, Corso Castelfidardo, 30/A, Turin, Italy
See maps:Main entrance and conference registration: https://www.google.it/maps/place/45%C2%B003'56.0%22N+7%C2%B039'37.6%22E/@45.0654608,7.660355,17.54z/data=!4m4!3m3!8m2!3d45.065545!4d7.660449?entry=ttu
Detailed Schedule:
2.00 - 2.30 Poster set up
2.30 - 2.40 Workshop opening: Welcome and introduction (by workshop organizers)
2.40 - 3.10 Invited talk 1 - Elisa Ricci
3.10 - 3.40 Invited talk 2 - Riccardo Volpi
3.40 - 3.50 Flash talks
3.50 - 4.00 Poster session (start)
4.00 - 4.30 Grab coffee + poster session
4.30 - 4.50 Poster session (continues)
4.50 - 5.20 Invited talk 3 - Oriane Siméoni
5.20 - 5.50 Invited talk 4 - Efstratios Gavves
5.50 - 6.00 Closing remarks and best paper awards
6.00 - 8.30 ECML-PKDD Welcome Reception @ Officine Grandi Riparazioni (Fucine)
(25min Talks + 5min Q&A)
Accepted Papers
The following accepted papers will be presented during the poster session:
Harnessing Error Patterns to Estimate Out-Of-Distribution Performance
Thomas BONNIER (SOCIETE GENERALE)*; Benjamin BOSCH (SOCIETE GENERALE)
Trajectory Prediction for Autonomous Safe Driving
Krishna Sri Ipsit Mantri (Purdue University)*; Nevasini NA Sasikumar (PESU)
HAVE-Net: Hallucinated Audio-Visual Embeddings for Few-Shot Classification with Unimodal Cues
Ankit Jha (Indian Institute of Technology Bombay)*; Debabrata Pal (Honeywell); Mainak Singha (Indian Institute of Technology Bombay); Naman Agarwal (IIT Bombay); Biplab Banerjee (Indian Institute of Technology, Bombay)
Efficient Cross-Domain Federated Learning by MixStyle Approximation
Manuel Röder (Technical University of Applied Sciences Würzburg-Schweinfurt)*; Leon Heller (Technical University of Applied Sciences Würzburg-Schweinfurt); Maximilian Münch (University of Groningen); Frank-Michael Schleif (THWS)
Monitoring under Varying Conditions: A Physics-Inspired Continuous Domain Adaptation
Maxime Leiber (Safran Tech)*; Yosra Marnissi (SAFRAN TECH); Dohy Hong (Safran Tech)
Key Design Choices for Double-Transfer in Source-Free Unsupervised Domain Adaptation
Andrea Maracani (Istituto Italiano di Tecnologia)*; Raffaello Camoriano (Politecnico di Torino); Elisa Maiettini (Istituto Italiano di Tecnologia); Davide Talon (Istituto Italiano di Tecnologia (IIT)); Lorenzo Rosasco (unige, mit, iit); Lorenzo Natale (Italian Institute of Technology)
CAD Models to Real-World Images: A Practical Approach to Unsupervised Domain Adaptation in Industrial Object Classification
Dennis Ritter (Berlin University of Applied Sciences); Mike Hemberger (nyris GmbH); Marc Hönig (topex GmbH); Volker Stopp (topex GmbH); Erik Rodner (University of Applied Sciences, Berlin); Kristian Hildebrand (Berlin University of Applied Sciences)*
EMG subspace alignment and visualization for cross-subject hand gesture classification
Martin Colot (Université Libre de Bruxelles)*; Gianluca Bontempi (Université Libre de Bruxelles); Cédric Simar (Université Libre de Bruxelles); Mathieu Petieau (Université Libre de Bruxelles); Ana Maria Cebolla Alvarez (Université Libre de Bruxelles); Guy Cheron (Université Libre de Bruxelles)
Adapting Classifiers To Changing Class Priors During Deployment
Abhijit Mahalanobis (Center for Research in Computer Vision, UCF); Natnael T Daba (University of Arizona)*; Bruce W McIntosh (UCF CRCV); Bruce Mcintosh (UCF CRCV)
Domain Generalization for Crop Segmentation with Knowledge Distillation
Simone Angarano (Politecnico di Torino)*; Mauro Martini (Politecnico di Torino); Alessandro Navone (Politecnico di Torino); Marcello Chiaberge (Politecnico di torino)
CoMFormer: Continual Learning in Semantic and Panoptic Segmentation
Fabio Cermelli (Politecnico di Torino)*; Matthieu Cord (Sorbonne University); Arthur Douillard (Heuritech / Sorbonne University)