4 papers
DiffusionBench: On Holistic Evaluation of Diffusion Transformers
Xingjian Leng, Jaskirat Singh, Zhanhao Liang +5
Diffusion transformer (DiT) research on image generation has converged to a single evaluation setup: class-conditional generation on ImageNet. While methods improve the FID and rel…
LeapAlign: Post-Training Flow Matching Models at Any Generation Step by Building Two-Step Trajectories
Zhanhao Liang, Tao Yang, Jie Wu +2
This paper focuses on the alignment of flow matching models with human preferences. A promising way is fine-tuning by directly backpropagating reward gradients through the differen…
Aesthetic Post-Training Diffusion Models from Generic Preferences with Step-by-step Preference Optimization
Zhanhao Liang, Yuhui Yuan, Shuyang Gu +5
Generating visually appealing images is fundamental to modern text-to-image generation models. A potential solution to better aesthetics is direct preference optimization (DPO), wh…
Glyph-ByT5: A Customized Text Encoder for Accurate Visual Text Rendering
Zeyu Liu, Weicong Liang, Zhanhao Liang +4
Visual text rendering poses a fundamental challenge for contemporary text-to-image generation models, with the core problem lying in text encoder deficiencies. To achieve accurate…