Recent advancements in autonomous driving are increasingly focused on enhancing safety and efficiency through innovative frameworks and technologies. The integration of natural language processing with dynamic maps is enabling more intuitive human-vehicle interactions, while novel transformer architectures are streamlining end-to-end driving systems by compressing multi-camera data without sacrificing accuracy. Additionally, diffusion models are being harnessed for motion planning, offering robust solutions to safety challenges in uncertain environments. Techniques such as risk-aware world model predictive control are addressing the limitations of traditional imitation learning by allowing vehicles to make informed decisions without relying solely on expert demonstrations. Furthermore, the development of anomaly detection frameworks is improving the identification of rare, high-risk scenarios, crucial for Level 4 autonomous vehicles. Collectively, these efforts are paving the way for more reliable and adaptable autonomous driving systems, with significant implications for urban mobility and transportation safety.
Top papers
- Driving on Registers(8.0)
- Talk2DM: Enabling Natural Language Querying and Commonsense Reasoning for Vehicle-Road-Cloud Integrated Dynamic Maps with Large Language Models(8.0)
- K-Gen: A Multimodal Language-Conditioned Approach for Interpretable Keypoint-Guided Trajectory Generation(8.0)
- SToRM: Supervised Token Reduction for Multi-modal LLMs toward efficient end-to-end autonomous driving(7.0)
- IRL-DAL: Safe and Adaptive Trajectory Planning for Autonomous Driving via Energy-Guided Diffusion Models(7.0)
- CycleBEV: Regularizing View Transformation Networks via View Cycle Consistency for Bird's-Eye-View Semantic Segmentation(7.0)
- DualShield: Safe Model Predictive Diffusion via Reachability Analysis for Interactive Autonomous Driving(7.0)
- PRAM-R: A Perception-Reasoning-Action-Memory Framework with LLM-Guided Modality Routing for Adaptive Autonomous Driving(7.0)
- SaFeR: Safety-Critical Scenario Generation for Autonomous Driving Test via Feasibility-Constrained Token Resampling(6.0)
- Natural Language Instructions for Scene-Responsive Human-in-the-Loop Motion Planning in Autonomous Driving using Vision-Language-Action Models(6.0)
- TaCarla: A comprehensive benchmarking dataset for end-to-end autonomous driving(6.0)
- Conditional Flow Matching for Continuous Anomaly Detection in Autonomous Driving on a Manifold-Aware Spectral Space(6.0)
- DrivePTS: A Progressive Learning Framework with Textual and Structural Enhancement for Driving Scene Generation(6.0)
- NoRD: A Data-Efficient Vision-Language-Action Model that Drives without Reasoning(5.0)
- Monte-Carlo Tree Search with Neural Network Guidance for Lane-Free Autonomous Driving(3.0)