Junhai Yong

Achieve photorealistic, identity-consistent facial video edits from text prompts without video training data, rivaling traditional rendering software.

Achieve significant speedups in diffusion model inference, without training, by adaptively selecting the best predictor for each token at each step based on a low-cost probe of the first layer.

Domain-specific knowledge hypergraphs can now be extracted with significantly improved quality by dynamically learning and applying extraction skills, outperforming static few-shot learning.