New AI Model Cuts Cost of Autonomous Driving by Ditching Expensive Sensors
Researchers have developed PRIX, an autonomous driving system that relies only on camera data rather than expensive LiDAR sensors, while running on smaller, faster AI models. The breakthrough could make self-driving technology practical for mass-market vehicles and accelerate commercial deployment by reducing both hardware costs and computational demands.
Originaltitel: PRIX: Learning to Plan From Raw Pixels for End-to-End Autonomous Driving
While end-to-end autonomous driving models show promising results, their practical deployment is often hindered by large model sizes, a reliance on expensive LiDAR sensors and computationally intensive BEV feature representations. This limits their scalability, especially for mass-market vehicles equipped only with cameras. To address these challenges, we propose <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">PRIX</b> (<bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">P</b>lan from <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">R</b>aw p<bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">IX</b>els). Our novel and efficient end-to-end driving architecture operates using only camera data, without explicit BEV representation and forgoing the need for LiDAR. PRIX leverages a visual feature extractor coupled with a generative planning head to predict safe trajectories from raw pixel inputs directly. A core component of our architecture is the Context-aware Recalibration Transformer (CaRT), a novel module designed to effectively enhance multi-level visual features for more robust planning. PRIX achieves SOTA performance on the NavSim-v2 and nuScenes datasets. On NavSim-v1, it also outperforms the majority of multimodal planners and other camera-only approaches on most metrics. Critically, PRIX is significantly more efficient on NavSim-v1, boasting faster inference speeds and a smaller model size. This combination of performance and efficiency makes it a practical solution for real-world deployment. Our work is open-source and the code will be available upon publication. Check our project website for more <uri xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">https://maxiuw.github.io/prix</uri>.