Yingpeng Zhang

Papers

Total Citations

H-Index

About

Yingpeng Zhang is a rising star in the field of autonomous driving, with a primary focus on end-to-end driving systems and generative robotic policy learning. His most notable contribution is the development of **DiffusionDrive**, a truncated diffusion model that dramatically accelerates action generation for real-time autonomous driving. By reducing the standard 100 denoising steps to just 4–8, Zhang’s work bridges the gap between the high-quality multi-modal action distributions offered by diffusion models and the stringent latency requirements of self-driving vehicles. This innovation has already garnered over 40 citations since its 2025 release, signaling strong impact in a rapidly evolving field. Zhang’s research addresses a critical bottleneck: how to leverage powerful generative techniques without sacrificing inference speed. His work is particularly relevant for students and researchers exploring diffusion-based policy learning, as it provides a practical framework for deploying these models in latency-sensitive robotics applications. With DiffusionDrive, Zhang has established himself as a key contributor to the next generation of autonomous driving architectures, where generative AI meets real-world deployment constraints.