Wenhan Luo

A novel method for creating detailed 3D images from single images using multiview diffusion.

CoCoGesture creates lifelike gestures that match spoken words, enhancing interaction.

New methods enhance main task performance using auxiliary data without extra computation costs.

MMTrail combines visual and audio descriptions for better video-language models.

A new method offers improved 3D modeling from just one image, enhancing realism.

Researchers develop new model for lively singing videos, enhancing animations.

A new framework shines light on hidden features of machine learning models.