4 papers across 3 sessions
We propose a new guidance method for flow based models for better sampling and generation quality.
We introduce LoRAShop, the first framework for multi-concept image generation and editing with LoRA models.
We introduce Collaborative Direct Preference Optimization (C-DPO) - a novel method that aligns image edits with user-specific preferences while leveraging collaborative signals from like-minded individuals
We introduce the first work for disentangled creative editing with an agentic framework.