Xinyu Tian

Existing video quality metrics fall short for camera-controlled generation, but CWQA sets a new standard by accurately predicting perceptual quality with a tailored approach.

Radiology data can be de-identified for cross-hospital sharing without sacrificing diagnostic utility, unlocking larger and more diverse training datasets for medical AI.

RL's success in boosting VLM reasoning hides a critical flaw: it crushes the model's ability to explore diverse solutions, leading to premature convergence and hindering scalability.

Ditch slow diffusion models: MAGT offers faster, single-pass generative sampling with better support fidelity by learning a manifold-aligned transport.