Let's use slope charts to compare points trend of the current Premier League to the last 20 seasons. Wolves, Liverpool FC and Leicester City are making one of their best starts while Tottenham Hotspur, Manchester United and Arsenal are having one of their worst. #dataviz Romain Vuillemot

Demo day for #dataviz student projects at Centrale Lyon w Théo Jaunet. Still amazed by students creativity and D3.js tech skills acquired in 6 weeks. Links to project websites soon..

Excited to attend #SSAC20 this week in Boston! We’ll present a characterization of space & time-dependence of 3-point shots in #Basketball. It's the result of a joint work at Centrale Lyon with Romain Vuillemot Romain Vuillemot (LIRIS), Wouter Bos (LMFA) and Nathan Rivière.

Considering the distance to a point (a), area inside green circle is controlled by the red team. However, player 1 will reach this area before the player 2 since the velocity of 1 is directed towards this one while 2 is moving away from it. Considering inertia (b) corrects this.

Happy to announce that a preprint of our ECCV 2020 spotlight paper "Learning to plan with uncertain topological maps" is now on Arxiv arxiv.org/abs/2007.05270 . We approximate a classical path planning algorithm and learn to plan under uncertainty. Christian Wolf (🦋🦋🦋) Chroma INRIA

In order to maintain collective stress level, I wrote a Twitter bot which daily tweets about the number of days left to the next ML & CV conference deadlines.

New video by CorentK, where he presents how we use weakly supervised word-object alignment to improve vision and language tasks. Joint work with Grigory Antipov and Moez Baccouche Video: youtube.com/watch?v=SW8J0G… Paper: arxiv.org/abs/1912.03063 (ECAI 2020)

Congratulations to Dr. Valentin Lachand Valentin Lachand Pascal ⏚ for a great remote defense and many thanks to the jury !👏🍾

Wow, one of our #CVPR2021 papers had a reviewer update his/her score from borderline to strong accept ... that was apparently a good job by CorentK Théo Jaunet Grigory Antipov Moez Baccouche Romain Vuillemot ... thanks to the reviewers and ACs!

3 articles acceptés à la conférence CVPR par les équipes IMAGINE+SICAL liris.cnrs.fr/actualite/3-ar…

Our #CVPR2021 paper is on arxiv: "How Transferrable are Reasoning Patterns in VQA?", by CorentK, Théo Jaunet, Grigory Antipov , Moez Baccouche, Romain Vuillemot. A deep analysis of attention in visual reasoning. arXiv: arxiv.org/abs/2104.03656 interactive visualization: reasoningpatterns.github.io

Congrats Théo Jaunet for his acceptance at the #NeuriPS2021 workshop on XAI approaches (...). He introduces a tool for visualizing the Sim2Real gap in robot navigation. With Guillaume Bono, Romain Vuillemot, yours truly. arxiv.org/abs/2109.11801 Tweeted after acceptance only (CVPR Motion)😉

Happening soon! Théo Jaunet will present his latest work at #ieeevis during the Mitigating Bias session (room frenchmen) joint work with CorentK Grigory Antipov Moez Baccouche Christian Wolf (🦋🦋🦋) & yours truly at LIRIS

For the 5th time, we're back! The Workshop on Visualization for AI Explainability (#visxai) will accept "explorables" & "explainables" that visualize, explain, and tell interactive stories about AI. #ieeevis ➡️ Get started early! Deadline: July 2022 visxai.io

Our #ICLR2022 paper has been accepted: "Filtered-CoPhy: Unsupervised Learning of Counterfactual Physics in Pixel Space". Lead by Steeven Janny (congrats!!), this was a collaboration between INSA Lyon Université Lyon 1 NAVER LABS Europe Simon Fraser University and AI at Meta. filteredcophy.github.io

PhD defense of Théo Jaunet on interpretibility of deep learning with Visual Analytics, right now. LIRIS INSA Lyon Romain Vuillemot

It has been quite a journey, but I just defended by PhD at LIRIS & INSA Lyon🥳! Thanks to Romain Vuillemot & Christian Wolf (🦋🦋🦋) for these amazing years! And thanks to the committee members: Céline Hudelot, David Auber, Shixia Liu, Hendrik S (find me in blueksy) & Liming Chen!