Research map / Image generation / Text-to-image diffusion
One Transformer Fits All Distributions in Multi-Modal Diffusion at Scale
Fan Bao, Shen Nie, Kaiwen Xue, Chongxuan Li, Shi Pu, Yaole Wang, Gang Yue, Yue Cao, Hang Su, Jun Zhu
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use one joint noise-prediction model with modality-specific timesteps instead of the multi-flow multi-task strategy | General-purpose image-text generation · MS-COCO validation, text-to-image generation | FID: 10.09 -> 9.71 | Improved |
| Use UniDiffuser instead of Versatile Diffusion | General-purpose image-text generation · Text-to-image and image-to-text generation | FID and CLIP score | Improved |
| Apply classifier-free guidance during conditional and joint sampling | UniDiffuser · Conditional and joint generation | Sample quality and image-text alignment | Improved |
| Use UniDiffuser instead of Versatile Diffusion | General-purpose image-text generation · Text-to-image generation, 10 samples and 25 denoising steps on one A100 80GB GPU | Inference time: 23.89s -> 19.77s | Improved |
| Use UniDiffuser instead of Versatile Diffusion | General-purpose image-text generation · Text-to-image generation, 10 samples and 25 denoising steps on one A100 80GB GPU | Inference memory: 76.53GB -> 48.30GB | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- GLIDE — GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models (ICML 2022) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Diffuse Everything: Multimodal Diffusion Models on Arbitrary State Spaces (ICML 2025 · similarity 0.91)
- Neural Diffusion Models (ICML 2024 · similarity 0.89)
- Dual Diffusion for Unified Image Generation and Understanding (CVPR 2025 · similarity 0.89)
- UniFL: Improve Latent Diffusion Model via Unified Feedback Learning (NeurIPS 2024 · similarity 0.89)
- Diffusion Models for Multi-Task Generative Modeling (ICLR 2024 · similarity 0.88)
- CleanDIFT: Diffusion Features without Noise (CVPR 2025 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
