arXiv subjects

Thibault Buhet

Publications and source records attributed to Thibault Buhet.

6 recordsLinked to original sources

arXiv2026

Pictura: Perspective-View Self-Play at Scale for Driving

Self-play in simulation produces robust driving policies at scale. Demonstrations of such behavior have been made using privileged vectorized observations such as exact poses and velocities, even for occluded agents. This assumes that perception is solved and introduces a representation gap with the partial observation of a deployed agent driving from the perspective view of egocentric cameras. A common fix, distilling the privileged policy into a camera-input student, leaves the student imitating decisions its own view cannot justify. Instead, we establish perspective-view self-play as a practical training regime. We introduce Pictura, a GPU-accelerated multi-agent driving simulator that renders each agent's egocentric view at every step, mitigating the representation gap at its source. Pictura sustains up to 500K agent-steps/s (2M images/s) on a single H100. Using Pictura, we train Alberti by self-play with plain PPO. It is the first large-scale driving self-play policy trained directly from perspective images, without privileged observations. Training spans 50B agent steps for ~35M km of driving. It approaches the driving performance of its privileged vectorized counterpart, and transfers zero-shot to Waymo Open Motion Dataset layouts re-rendered in Pictura, where it outperforms privileged vectorized agents. Project page: https://valeoai.github.io/Pictura/

cs.CV↗

arXiv2025

Multiple-Frequencies Population-Based Training

Reinforcement Learning's high sensitivity to hyperparameters is a source of instability and inefficiency, creating significant challenges for practitioners. Hyperparameter Optimization (HPO) algorithms have been developed to address this issue, among them Population-Based Training (PBT) stands out for its ability to generate hyperparameters schedules instead of fixed configurations. PBT trains a population of agents, each with its own hyperparameters, frequently ranking them and replacing the worst performers with mutations of the best agents. These intermediate selection steps can cause PBT to focus on short-term improvements, leading it to get stuck in local optima and eventually fall behind vanilla Random Search over longer timescales. This paper studies how this greediness issue is connected to the choice of evolution frequency, the rate at which the selection is done. We propose Multiple-Frequencies Population-Based Training (MF-PBT), a novel HPO algorithm that addresses greediness by employing sub-populations, each evolving at distinct frequencies. MF-PBT introduces a migration process to transfer information between sub-populations, with an asymmetric design to balance short and long-term optimization. Extensive experiments on the Brax suite demonstrate that MF-PBT improves sample efficiency and long-term performance, even without actually tuning hyperparameters.

cs.LG↗

arXiv2025

V-Max: A Reinforcement Learning Framework for Autonomous Driving

Learning-based decision-making has the potential to enable generalizable Autonomous Driving (AD) policies, reducing the engineering overhead of rule-based approaches. Imitation Learning (IL) remains the dominant paradigm, benefiting from large-scale human demonstration datasets, but it suffers from inherent limitations such as distribution shift and imitation gaps. Reinforcement Learning (RL) presents a promising alternative, yet its adoption in AD remains limited due to the lack of standardized and efficient research frameworks. To this end, we introduce V-Max, an open research framework providing all the necessary tools to make RL practical for AD. V-Max is built on Waymax, a hardware-accelerated AD simulator designed for large-scale experimentation. We extend it using ScenarioNet's approach, enabling the fast simulation of diverse AD datasets.

cs.LG↗

arXiv2020

PLOP: Probabilistic poLynomial Objects trajectory Planning for autonomous driving

To navigate safely in urban environments, an autonomous vehicle (ego vehicle) must understand and anticipate its surroundings, in particular the behavior and intents of other road users (neighbors). Most of the times, multiple decision choices are acceptable for all road users (e.g., turn right or left, or different ways of avoiding an obstacle), leading to a highly uncertain and multi-modal decision space. We focus here on predicting multiple feasible future trajectories for both ego vehicle and neighbors through a probabilistic framework. We rely on a conditional imitation learning algorithm, conditioned by a navigation command for the ego vehicle (e.g., "turn right"). Our model processes ego vehicle front-facing camera images and bird-eye view grid, computed from Lidar point clouds, with detections of past and present objects, in order to generate multiple trajectories for both ego vehicle and its neighbors. Our approach is computationally efficient and relies only on on-board sensors. We evaluate our method offline on the publicly available dataset nuScenes, achieving state-of-the-art performance, investigate the impact of our architecture choices on online simulated experiments and show preliminary insights for real vehicle control

cs.CV↗

arXiv2019

Conditional Vehicle Trajectories Prediction in CARLA Urban Environment

Imitation learning is becoming more and more successful for autonomous driving. End-to-end (raw signal to command) performs well on relatively simple tasks (lane keeping and navigation). Mid-to-mid (environment abstraction to mid-level trajectory representation) or direct perception (raw signal to performance) approaches strive to handle more complex, real life environment and tasks (e.g. complex intersection). In this work, we show that complex urban situations can be handled with raw signal input and mid-level representation. We build a hybrid end-to-mid approach predicting trajectories for neighbor vehicles and for the ego vehicle with a conditional navigation goal. We propose an original architecture inspired from social pooling LSTM taking low and mid level data as input and producing trajectories as polynomials of time. We introduce a label augmentation mechanism to get the level of generalization that is required to control a vehicle. The performance is evaluated on CARLA 0.8 benchmark, showing significant improvements over previously published state of the art.

cs.AI↗

arXiv2018

Imitation Learning for End to End Vehicle Longitudinal Control with Forward Camera

In this paper we present a complete study of an end-to-end imitation learning system for speed control of a real car, based on a neural network with a Long Short Term Memory (LSTM). To achieve robustness and generalization from expert demonstrations, we propose data augmentation and label augmentation that are relevant for imitation learning in longitudinal control context. Based on front camera image only, our system is able to correctly control the speed of a car in simulation environment, and in a real car on a challenging test track. The system also shows promising results in open road context.

cs.CV↗