Patrick Pérez’s page

Patrick Pérez is CEO at Kyutai, a non-profit open-science AI lab, based in Paris. Prior to this, Patrick was at Valeo as VP of AI and Scientific Director of valeo.ai (2018-2023), and with Technicolor (2009-2018), Inria (1993-2000, 2004-2009) and Microsoft Research Cambridge (2000-2004) as research scientist. His research interests lie in reliable multimodal AI for the benefit of all.

News

2024.11: Keynote at ai-Pulse (video)
2024.10: Keynote at AI & BigData Paris
2024.09: Paper accepted at NeurIPS’24 on 3D human pose estimation
2024.09: We release paper, inference codes and models for Moshi (paper, github)
2024.07: Kyutai introduces Moshi, first ever-released real-time voice AI (EU demo, US demo, press release, keynote); paper, weights and codes under preparation for release, stay tuned.
2024.07: Papers accepted at ECCV’24.
2024.07: Keynote and panel at the Hi!Paris Summer School.
2024.06: Panel at the Meta’s Open Source AI Day, at Station F.
2024.06: Survey paper on unsupervised object discovery accepted in IJCV, project led by Oriane Siméoni
2024.06: Panel at the Cyber Valley Days in Tübingen.
2024.06: Keynote at the delightful Imagine-ENPC retreat.
2024.05: Paper accepted at ICML’24.
2024.04: Papers accepted at CVPR’24.
2024.04: Panel at the R.AI.SE Summit 2024 (video and itw with BFMTV)
2023.12: Joining Kyutai as CEO.
2023.12: Paper on monocolar hand-face reconstrution at SiggraphAsia (collaboration with MPI).
2023.10: Podcast on automotive AI in Moving Your Future by Valeo series
2023.09: Participating to several events at ICCV in Paris: BRAVO workshop, Human Environment Navigation workshop, and tutorial on Reliability.
2023.09: Papers accepted at NeurIPS’23.
2023.07: Paper on prompt-based zero-shot domain adaptation accepted at ICCV 2023.
2023.07: Radio program at Rencontres de Pétraques (“Comment dépasser la guerre des intelligences”) by Caroline Broué, along with Anne Alombert, Yoshua Bengio and Yann LeCun.
2023.07: Talk on AI at CNIL (Commission Informatique et Liberté)
2023.06: Talk at Object localization for free: Going beyond self-supervised learning CVPR’23 Tutorial.
2023.06: Keynote at Vision4AllSeasons CVPR’23 workshop.
2023.05: Talk and panels at CIIRC 10-th anniversary’s AI Symposium
2023.05: Keynote at Orasis 2023.
2023.04: BRAVO: Robustness and Reliability of Autonomous Vehicles in the Open-world accepted as ICCV’23 workshop.
2023.04: “The Many Faces of Reliability of Deep Learning for Real-World Deployment” accepted as ICCV’23 tutorial.
2023.03: Papers accepted at CVPR’23.
2023.03: Valeo.ai celebrates its 5th anniversary!
2023.01: Paper on self-supervised learning with rotation-invariant kernels accepted at ICLR’23 (collab with ENS Lyon).
2022.12: Paper on transformer-based BEV segmentation preesented at CoRL’22 (collab with Inria Grenoble).
2022.11: Talk at Collège de France workshop on the Future of Transports (video).
2022.11: Article in Les Echos (in French).
2022.11: Talk at Wayve Scientific Speaker Series.
2022.10: Tutorial Self-supervision on wheels: Advances in self-supervised learning from autonomous driving data at ECCV’22 (video).
2022.10: Keynote at the VBSD ECCV workhsop
2022.10: Plenary talk at ICIP’22.
2022.10: Keynote at Confiance.ai Days in Saclay.
2022.09: Keynote, panel and chair at Czech-French workshop on AI in Prague.
2022.09: Talk at Andrew Zisserman Festschrift, in Oxford.
2022.07: Code for our ECCV’22 paper “Active learning strategies for weakly-supervised object detection”.
2022.07: Code for our ECCV’22 paper “Steering Counterfactual Explanations with Semantics”.
2022.07: Course on robustness, confidence and adaptation at the Vision and Sport Summer School (VS3 2022) in Prague.
2022.07: Our survey on the Explainability of vision-based autonomous driving systems accepted in IJCV.
2022.07: Papers accepted at ECCV’22.
2022.06: Keynote at the Vision4AllSeasons CVPR Workshop.
2022.05: Keynote at the Autonomy2.0 ICRA Workshop.
2022.02: Talk at EPFL CV Seminar.
2022.01: IMT podcast (in French) on challenges of driving automation.
2022.01: Keynote on trustworthy AI at the French Academy of Technologies.
2022.01: Keynote on basic AI research in corporate labs at Eurecom.
2021.11: Keynote at the ACM-Computer Science in Cars Symposium (CSCS).
2021.11: Panel on AI at Institut Henri Pointcaré.
2021.11: Panel on autonomous vehicles at the AI Day of the French AI Manifesto.
2021.10: Panel on Machine Learning and Physics at the Industry Day on AI of the French Physics Society.
2021.09: Paper accepted at NeurIPS’21 (collb with ENS/Inria-Paris and NYU).
2021.08: Paper on Detecting 32 pedestrian attributes for autonomous vehicles with code accepted in T-IST (collab with EPFL).
2021.08: Code for our ICCV’21 paper “Multi-view radar semantic segmentation”.
2021.08: Paper on Handling new target classes in semantic segmentation with domain adaptation accepted in CVIU.
2021.07: Papers accepted at ICCV’21.
2021.07: Code for our CVPR’21 paper “Semantic Palette: Guiding scene generation with class proportions”.
2021.06: Keynote at Vision4AllSeasons CVPR’21 workshop.
2021.06: Keynote at SAIAD CVPR’21 workshop.
2021.05: Paper on Confidence estimation via auxiliary models accepted in IEEE T-PAMI (collab with CNAM).
2021.04: Paper accepted at Siggraph’21 (collab with MPI).
2021.03: Panel on Math and AI.
2021.03: Papers accepted at CVPR’21.
2021.02: Industrial panel at Visigrapp 2021.
2021.02: Post on the explainability of neural driving models.
2020.12: Paper accepted at AAAI’20 (collab with CTU).
2020.12: Keynote at the Machine Learning for Autonomoud Driving (ML4AD) virtual NeurIPS workshop [video].
2020.11: Talk at NaverLabs Europe.
2020.10: Talk at Samsung AI Center (SAIC).
2020.09: Paper accepted at SiggraphAsia’20.
2020.07: Papers accepted at ECCV’20.
2020.06: Keynote at the Safe AI for Automated Driving (SAIAD) CVPR’20 workshop [video].
2020.06: Code for CVPR’20 paper “xMUDA: Cross-Modal Unsupervised Domain Adaptation for 3D Semantic Segmentation” (xMUDA).
2020.02: Papers accepted at CVPR’20.
2019.12: Code for ICCV’19 paper “Boosting few-shot visual learning with self-supervision” (BF3S).
2019.12: Code for ICCV’19 paper “DADA: Depth-Aware Domain Adaptation in semantic segmentation” (DADA).
2019.11: OmniCV workshop accepted at CVPR’20.
2019.10: Talk at Prairie AI Summer School (PAISS’19).
2019.10: Panel on future mobilities at PRAIRIE inauguration day.
2019.09: Code of NeurIPS’19 paper “Addressing failure prediction by learning model confidence” (valeo.ai ConfidNet).
2019.09: Papers accepted at NeurIPS’19.
2019.07: Papers accepted at ICCV’19.
2019.06: Code of CVPR’19 paper “Adversarial Entropy Minimization for Domain Adaptation in Semantic Segmentation” (AdvEnt).
2019.06: Code of CVPR’19 paper “Sorting Deep Net to learn ranking loss surrogates” (DeepSort).
2019.06: Keynote at the Safe AI for Automated Driving (SAIAD) CVPR’19 workshop.
2019.03: Papers accepted at CVPR’19.
2019.03: Talk at AI & Physics (Pascal Institute, Orsay).
2019.03: Talk at Imaging & Machine Learnning (Poincaré Institute, Paris).
2019.03: Talk at Computing and Learning (Lyon).
2018.11: Code of CPVR’18 paper “Finding beans in burgers: Deep semantic-visual embedding with localization” (DVSE-loc).
2018.11: Talks at SophI.A.2018 (Sophia Antipolis) and Horizon Maths 2018 : Intelligence Artificielle (ENS Paris).
2018.09: Workshop “Deep learning: From theory to applications”.
2018.08: Deep Video Portraits presented at Siggraph (statement,video).
2018.07: Course at “Summer School on Signal and Image Processing”, in Peyresq (lovely village in the French Alps).
2018.03: Start of Valeo.ai in the center of Paris.