Junho Lee

A new method enhances image generation by organizing latent space in diffusion models.

A new method improves video classification by optimizing frame selection.

A new training method improves image segmentation for specific object identification.