Xueyang Fu

World action models do not need to render clean future video to generalize: conditioning on a single forward pass of pure noise captures virtually all the robustness benefits of multi-step diffusion at a fraction of the compute.

Fusing event data with global illumination unlocks significantly better low-light image enhancement, outperforming prior art by up to 1.24dB PSNR and 0.069 SSIM.

Event cameras unlock robust gait recognition even in challenging low-light conditions, outperforming traditional cameras by separately modeling motion dynamics and static shape.

Linear transport flows between degraded and clean image domains enable fast, adaptable image restoration that outperforms existing methods in distortion-perception balance.

Forget slow, INR-based video super-resolution: 2D Gaussian Splatting now delivers state-of-the-art quality with a 3x speedup, even at extreme temporal scales.