Image-to-LoRA V2 compresses style into single forward pass, no training

Arxiv CS.CV (computer vision)

Updated i2L method turns one or more reference images into a style LoRA in a single forward pass, eliminating the need for explicit per-style training. The approach achieves efficient style transfer while preserving reference-specific appearance.

Diffusion modelsLoRA

See 1 more source

Zero training needed, Image-to-LoRA(i2L) V217 hours agoswitch2stockDiscuss

Arxiv CS.CV (computer vision)·Zhongjie DuanYingda Chen

2 days ago