About the ILLS - DATAIA Institute workshop
The collaborative workshop in artificial intelligence (AI) between ILLS and DATAIA aims to bring together experts from various French and Canadian academic and research institutions to share their knowledge and ideas on the latest trends and advancements in the field of AI. This workshop provides a platform for participants to present their research work, discuss their findings, and collaborate with peers to identify new research directions and opportunities.
The workshop covers a wide range of topics related to AI, including natural language processing, computer vision, machine learning, and deep learning. Participants will share their experiences and insights on the challenges and opportunities in these areas and discussed how they can be addressed through collaboration and interdisciplinary research. In addition to research presentations and discussions, the workshop also featured brainstorming sessions with renowned experts in the field who will share their perspectives on the future of AI and the role of collaboration between France - Canada in advancing the field.
Overall, this collaborative workshop provides a unique opportunity for participants to engage in meaningful discussions, to foster a community of researchers to build new collaborations, and to explore new avenues for research in AI.
Registration free but mandatory. Coffee breaks will be available for the participants (lunch not included).
About the IRL ILLS Center
The IRL ILLS centre gathers together McGill University, École de technologie supérieure (ÉTS), Mila – Quebec AI Institute, France’s Centre Nationale de la Recherche Scientifique (CNRS), Université Paris-Saclay, and the École CentraleSupéle. It emphasizes interdisciplinary collaborations with an aim to develop new methodologies and integrate these techniques into learning systems.
Wednesday, May 24th
Venue: Amphi sc.046 (Peugeot), Bouygues Building, CentraleSupélec
- 8:30 - 9am : welcome coffee
9 - 10:30am: 1st session "AI for Personalized Medicine"
Keywords: Computer vision, Medical image analysis, Deep learning
- 9am: Tal Arbel (Mila - ILLS, McGill) - The promise of AI for image-based personalized medicine
- 9:45am: Maria Vakalopoulou (MICS, CentraleSupélec, DATAIA) - Deep learning in medical imaging and digital pathology
- 10:05am: Nikos Paragios (TheraPanacea) - What if we could significantly improve cancer care by making precision medicine a reality?
- 10:30 - 11am: pause
11am - 12:30pm: 2nd session "Robust learning and unlabeled data"
Keywords: Computer vision, machine learning, image analysis
- 11am: Christian Desrosiers (Livia - ILLS, ETS) - Image analysis with unlabeled data: Doing more with less
- 11:45am: Nicolas Thome (UPMC, Sorbonne University) - Robust and hybrid machine learning
- 12:10pm: Florent Bouchard (L2S, CNRS) - Exploiting robust statistics for EEG classification
- 12:30 - 2:30pm: lunch break
2:30 - 3:35pm: 3rd session "Societal impacts of AI solutions"
Keywords: Roboethics, AI ethics, Societal impacts of AI solutions
- 2:30pm: AJung Moon (Mila - ILLS, McGill) - Responsible design of interactive systems
- 3:15pm: Claire Monteleoni (LISN, Inria) - Research on machine learning for climate change and environmental sustainability
- 3:35pm: Sonia Ben-Mokhtar (LIRIS, CNRS) - Decentralized learning (as a facilitator) for decentralized online services
Thursday, May 25th
Venue: Amphi II, Eiffel Building, CentraleSupélec
- 8:30 - 9am : welcome coffee
9 - 10:30am: 1st session "Intelligent Systems and Conversational Agents"
Keywords: Human-Computer Interaction, Robots
- 9am: Jeremy R. Cooperstock (CIM - ILLS, McGill) - Intelligent Systems to Enhance Human Experience
- 9:45am: Xavier Alameda-Pineda (RobotLearn Team, Inria) - Learning for Robots in Conversational Groups
- 10:05am: Caroline Appert (LISN, CNRS) - Tangible and Bare Hand Input for Interacting with Various Types of Displays
- 10:30 - 11am: break
11am - 12:30pm: 2nd session "Linguistic models, generation and understanding"
Keywords: Natural Language Models, Applications
- 11am: François Yvon (Sorbonne University, CNRS) - Wonders and mysteries of multilingual language models
- 11:20am: Nona Naderi (LISN, Université Paris-Saclay) - Natural language processing for public health and clinical science
- 11:40am: Frédéric Bechet (LIS, Aix Marseille University) - Natural language understanding: From classification to generation and back again?
- 12:05pm: Maks Ovsjanikov (Lix, IPP) - Efficient and robust learning on non-rigid surfaces and graphs
- 12:30 - 13:30pm: lunch break
1:30 - 3pm: 3rd session "Few-Shot and Learning and weakly supervised learning"
Keywords: Machine learning, Computer vision
- 1:30pm: Eric Granger (Livia - ILLS, ETS) - Adapting deep neural networks for video recognition with weakly labeled data
- 2:10pm: Jose Dolz (Livia - ILLS, ETS) - Learning with limited supervision
- 3:45 - 5pm: DATAIA Seminar - Jonathan Niles-Weed
Friday, May 26th
Venue: Amphi sc.046 (Peugeot), Bouygues Building, CentraleSupélec
- 8:30 - 9am : welcome coffee
9 - 10:30am: 1st session "Optimization in Machine Learning"
Keywords: Optimization in ML, Learning theory
- 9am: Marco Pedersoli (Livia - ILLS, ETS) - Learning is a two-level optimization
- 9:45am: Emilie Chouzenoux (CVN, Inria, DATAIA) - PrimAl Dual Minimum Description LEngth (PADDLE) for practical learning with a few moves
- 10:05am: Gabriele Facciolo (Centre Borelli, ENS Saclay) - Self-supervised image restoration
- 10:30 - 11am: break
11am - 12:30pm: 2nd session "Learning on graphs and decentralized learning"
Keywords: Computer networks, Learning on graphs, Federated learning
- 11am: Mark Coates (CIM - ILLS, McGill) - Learning from data on graphs
- 11:45am: Laurent Oudre (Centre Borelli, ENS Saclay, DATAIA) - Graphical signal processing for the study of multivariate physiological signals
- 12:30 - 2pm: lunch break
2 - 3:30pm: 3rd session "AI Solutions for Video Challenges and Environmental Impacts"
Keywords: Signal Processing, Deep Learning, Compression, Video Processing
- 2pm: Stéphane Coulombe (ILLS, ETS) - Immersive video challenges and some AI solutions
- 2:45pm: Anne-Laure Ligozat (LISN, Université Paris-Saclay) - Environmental impacts of AI
- 3:05pm: Giuseppe Valenzise (L2S, CNRS) - AI-based compression for immersive video: Current trends and future directions
- 3:30 - 4pm: break
4 - 4:50pm: 4th session "Online learning"
Keywords: Reinforcement learning, Bandits, Learning theory
- 4pm: Hédi Hadiji (L2S, CentraleSupélec) - Adaptation to unknown parameters in online learning and bandits
- 4:25pm: Silviu Maniu (LISN, Université Paris-Saclay) - Multi-armed bandits for influence propagation
To attend the remote seminar (via Teams):
https://www.microsoft.com/fr-fr/microsoft-teams/join-a-meeting?rtc=1
Meeting ID: 326 746 607 654
Secret Code: P8KPH4
Dates: from Wednesday 24 to Friday 26 May 2023.
Locations (CentraleSupélec, Gif-sur-Yvette) :
- Wednesday, May 24: Amphi sc.046 (Peugeot), Bouygues Building
- Thursday May 25: Amphi II, Eiffel Building
- Friday, May 26: Amphi sc.046 (Peugeot), Bouygues Building
Access to reach the workshop from Paris: RER B - stop "Le Guichet" then take bus 9 towards Campus HEC, Jouy-en-Josas and get off at "Moulon" stop.
Closing date for registration : May 22th, 2023