DreamCache: Finetuning-Free Lightweight Personalized Image Generation via Feature Cachingby Emanuele Aiello, Umberto Michieli, Diego Valsesia,…
Reward Fine-Tuning Two-Step Diffusion Models via Learning Differentiable Latent-Space Surrogate Rewardby Zhiwei Jia, Yuesong Nan,…
Efficient Pruning of Text-to-Image Models: Insights from Pruning Stable Diffusionby Samarth N Ramesh, Zhixue ZhaoFirst…
HeadRouter: A Training-free Image Editing Framework for MM-DiTs by Adaptively Routing Attention Headsby Yu Xu,…
Preference Alignment for Diffusion Model via Explicit Denoised Distribution Estimationby Dingyuan Shi, Yong Wang, Hangyu…
ComfyGI: Automatic Improvement of Image Generation Workflowsby Dominik Sobania, Martin Briesch, Franz RothlaufFirst submitted to…
Edge-Cloud Routing for Text-to-Image Model with Token-Level Multi-Metric Predictionby Zewei Xin, Qinya Li, Chaoyue Niu,…
Constant Rate Schedule: Constant-Rate Distributional Change for Efficient Training and Sampling in Diffusion Modelsby Shuntaro…
Generating Compositional Scenes via Text-to-image RGBA Instance Generationby Alessandro Fontanella, Petru-Daniel Tudosiu, Yongxin Yang, Shifeng…
SmoothCache: A Universal Inference Acceleration Technique for Diffusion Transformersby Joseph Liu, Joshua Geddes, Ziyu Guo,…