Xinbang Zhang

Papers

Total Citations

H-Index

About

Xinbang Zhang is a rising star in autonomous driving and generative robotics, whose work bridges cutting-edge diffusion models with real-world driving systems. His most-cited paper, "DiffusionDrive: Truncated Diffusion Model for End-to-End Autonomous Driving" (2025, 40 citations), introduces a novel approach that dramatically reduces the computational burden of diffusion-based policy learning—a key bottleneck in deploying these models for real-time driving. By proposing a truncated denoising strategy, Zhang demonstrates how to efficiently capture multi-modal action distributions without the heavy iterative sampling typical of robotic diffusion models. This work has already garnered attention for its potential to make end-to-end autonomous driving both safer and more scalable. Beyond this, Zhang’s research focuses on integrating generative AI with control systems, aiming to replace traditional modular pipelines with unified, learned policies. His contributions are particularly notable for addressing the trade-off between model expressiveness and inference speed, a critical challenge in robotics. With a rapidly growing citation record and a forward-looking approach, Xinbang Zhang is establishing himself as a key innovator at the intersection of generative modeling and autonomous systems.