Florian Fuchs

A new release of our RL agent GT Sophy as a permanent feature of GT7 controlling 340 car types on 9 tracks 🏎️💨

Job opening🚨 Tired of AI agents limited to coding and web browsing? I am looking for a PhD student (overseas or UK) to work with @j_foerst at @UniofOxford and myself at @iconicgamesio, full details below deadline is 29th of Jan!

Exited to share that the work I did at @SonyAI_global is public now and will be presented at @NeurIPSConf! How fun would be to have a single reinforcement learning agent providing diverse competitive policies in AAA games? Check out ai.sony/publications/D…! Fast🧵(1/7)

Reinforcement learning in #AI is hard, so I’ve made a website to collect answers I’ve given to common RL questions. It's named Decisions & Dragons. It’s launching with 8 questions and answers, but I will add to it in the future. A 🧵to give a preview with the link below.

Sony AI wins "Best Overall Use of AI in Gaming" from the 2024 AI Breakthrough Awards! GT Sophy 2.0 brings advanced AI to Gran Turismo 7, offering a thrilling racing experience. Check out our press release here bit.ly/3VBBLda #SonyAI #AIBreakthrough #GTSophy #GranTurismo7

@davsca1 @Nature Congrats!!

We are thrilled to share our groundbreaking paper published today in @Nature: "Low Latency Automotive Vision with Event Cameras." Paper: nature.com/articles/s4158… Video: youtu.be/dwzGhMQCc4Y Code & Dataset: github.com/uzh-rpg/dagr Frame-based sensors such as the RGB cameras used in the automotive industry face a bandwidth–latency trade-off: higher frame rates reduce perceptual latency but increase bandwidth demands, whereas lower frame rates save bandwidth at the cost of missing vital scene dynamics due to increased perceptual latency (see Fig. 1a of the paper). Event cameras have emerged as alternative vision sensors to address this trade-off. Event cameras measure the changes in intensity asynchronously, offering high temporal resolution and sparsity, markedly reducing bandwidth and latency requirements. Despite these advantages, event-camera-based algorithms are either highly efficient but lag behind image-based ones in accuracy or sacrifice the sparsity and efficiency of events to achieve comparable results. To overcome this, we propose a hybrid event- and frame-based object detector based on Deep Asynchronous GNNs, which preserves the advantages of each modality and thus does not suffer from this trade-off. Our method exploits the high temporal resolution and sparsity of events and the rich but low temporal resolution information in standard images to generate efficient, high-rate object detections, reducing perceptual and computational latency. In doing so, it emulates the slow-fast pathways in biological neural networks and uses them to its advantage. We show that using a 20-Hz RGB camera plus an event camera achieves the same latency as a 5,000-Hz camera with the bandwidth of a 50-Hz camera, i.e., an over 100-fold bandwidth reduction, without compromising accuracy. Our approach paves the way for efficient and robust perception in edge-case scenarios by uncovering the potential of event cameras. We release the code and the dataset (DSEC-Detection) to the public. Kudos to @DanielGehrig6, who, with this work, also received the UZH Annual Award for the Best PhD thesis! **Reference** Daniel Gehrig, Davide Scaramuzza Low Latency Automotive Vision with Event Cameras Nature, May 29, 2024. DOI: 10.1038/s41586-024-07409-w PDF (Open Access): nature.com/articles/s4158… Video (Narrated): youtu.be/dwzGhMQCc4Y Code & Datasets: github.com/uzh-rpg/dagr @UZH_en @UZH_Science @UZHspacehub @ERC_Research @nccrrobotics

Congratulations to the whole Sony AI team on what is surely one of the largest ever commercial deployments of an end-to-end reinforcement learning agent! It’s great to see all the positive feedback from gamers!

RL PhD students: Want to work on hard RL problems with applications to modern video games? Apply for an internship position at Sony AI! #ReinforcementLearning #AI ai.sony/joinus/job-rol…

@omiguelnavida Thanks <3

A new release of our RL agent GT Sophy as a permanent feature of GT7 controlling 340 car types on 9 tracks 🏎️💨

@antoniloq Congrats!

Excited to share that I'll join UPenn as an Assistant Professor next fall! I couldn't be more grateful to all the mentors and collaborators for their support along the way! I'm looking for motivated students👩‍🎓to join my lab! If you love robots 🤖 and learning, look no further!

Sony’s Gran Turismo AI racer can drift now, making it even more unbeatable engt.co/45IA6Xf

Was a lot of fun to work on this unusal application of RL 💨

@egrefen @MinqiJiang @_rockt @UCL_DARK Congrats @MinqiJiang 🥳

Congratulations Dr @MinqiJiang! @_rockt and I are so proud of the first PhD graduate from @UCL_DARK 🥰

You can find @Florian_fuu's @ic_arl presentation on how #GTSophy was trained using novel deep reinforcement learning techniques to outrace champion drivers in @thegranturismo below.

Many thanks to @Florian_fuu for the presentation at the ICARL seminar series! 🙌 Many valuable insights on Reinforcement Learning for Game AI 🏎️ If you’ve missed it, you may find the recording on YouTube 🎥 youtu.be/arkNKrMR66w

Very excited to present our work on Reinforcement Learning for Game AI at the Imperial College London! 🚗 I really appreciate that the organizers also allow everyone interested to join remotely

Scholarships available to participate in the AI and Games Summer School, generally sponsored by Sony! In particular aimed at people who could not otherwise participate. Apply now!

We had @robertarail, a research scientist at FAIR @MetaAI who is working on building generally capable autonomous agents at last year's summer school. Watch this exclusive interview with @sysnet_ai from @modl_ai and learn more about her here: youtu.be/n8UXLWNaMQ4 #airesearch

#GTSophy and 'Race Together' wouldn't have been possible without the teams' ingenuity and dedication. Thank you to PDI and Sony Interactive Entertainment Cloud Gaming teams for collaborating with Sony AI to make this project an incredible breakthrough success.

@MinqiJiang @_samvelyan Great work & thanks for the pointer! It would indeed be interesting to try that.

@_samvelyan @Florian_fuu check this out! Maybe relevant for a more generalist Sophy?

I’m excited to share our latest #ICLR2023 paper 🏎️ MAESTRO: Open-Ended Environment Design for Multi-Agent Reinforcement Learning 🏎 Paper: arxiv.org/abs/2303.03376 Website: maestro.samvelyan.com Highlights: 👇