Hao-Tian Yang

Adversarial distillation achieves state-of-the-art few-step visual generation by learning log-density ratios directly via classification, eliminating the need for auxiliary score fitting.

Enforcing uniform load balancing across video tokens fundamentally breaks visual coherence, but bifurcating experts into semantic and generic tiers resolves patch fragmentation while driving emergent coarse-to-fine denoising.

Severe oversaturation and artifacts in few-step video distillation stem directly from compounding critic errors, which can be neutralized via a single-line geometric projection without discarding the true generative signal.

DAC-Pose achieves remarkable fidelity in human generation, maintaining texture and identity consistency even under extreme viewpoint changes.

GAFIC achieves unprecedented accuracy in image cropping by fusing local and global features, outperforming traditional methods that compromise pixel integrity.