Yiyi Zhang

LaST outperforms traditional models by harnessing the strengths of both large and small architectures, achieving unprecedented accuracy in zero-shot surgical phase recognition.

Dynamic integration of multimodal knowledge boosts the recognition of rare surgical actions, overcoming critical optimization conflicts in robot-assisted surgery.

Reconstruct a high-fidelity, full-head 3D avatar from a single image in under one second, finally breaking the quality-speed tradeoff.

Stop training all your data the same way: this new method adaptively reweights samples based on how "competitive" they are within a group, leading to better performance and robustness.

MLLMs can achieve state-of-the-art multimodal retrieval by learning to compress information into a handful of "bottleneck" tokens, forcing the model to distill relevant semantics.