Tongliang Liu

A new method in machine learning enhances model adaptability across various data types.

A new approach to image segmentation enhances object recognition across various sectors.

Integrating language models into causal discovery streamlines analysis and reveals hidden relationships.

EchoAlign modifies data features to align with noisy labels, improving machine learning performance.

A new method enhances image perception in language models using diffusion models.

Introducing PART, a method to boost machine learning models' accuracy and robustness.

A new method improves causal discovery accuracy using automated kernel selection.

A framework to secure image privacy while maintaining model accuracy.

A new method enhances detection of unexpected data in machine learning models.

LaVin-DiT enhances how machines perceive and interpret visual data.

FedGR improves federated learning by refining noisy labels for better collaboration.

Urban4D redefines urban scene reconstruction for smarter cities.

A fresh method to identify AI-created images using predictive uncertainty.