AI Paper Atlas

Research map / Image generation / Text-to-image diffusion

Muse: Text-To-Image Generation via Masked Generative Transformers

Huiwen Chang, Han Zhang, Jarred Barber, Aaron Maschinot, Jose Lezama, Lu Jiang, Ming-Hsuan Yang, Kevin Patrick Murphy, William T. Freeman, Michael Rubinstein, Yuanzhen Li, Dilip Krishnan

ICML 2023 · Image generation › Text-to-image diffusion · approach: Personalized and region-controlled generation · 117 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

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 changedWhereMeasuredEffect
Use of parallel iterative decoding with cosine schedule (24 steps base, 8 steps super-res) instead of per-token autoregressive decodingMuse inference · Inference speed (TPUv4)per-batch inference time: Muse-3B 256x256 0.5s vs Parti-3B 6.4sImproved
Super-resolution stage added on top of base model (base + super-res vs base alone) (ablation)Muse on CC3M · CC3MFID: 6.8 -> 6.06Improved
Scaling Muse to 3B parameters with T5-XXL conditioning, zero-shot on COCOMuse · MS-COCO zero-shot 256x256FID-30K: 7.88 vs Parti-20B 7.23 and Imagen 7.27Mixed
Frozen pre-trained T5-XXL LLM text embeddings as conditioning (vs learning language model from scratch) (ablation)Muse · Qualitative/initial experimentsperformanceImproved
Linearly increasing classifier-free guidance scale during iterative samplingMuse sampling · CFG trade-offdiversity vs fidelityNo clear effect
Decoder finetuning with more residual blocks and channels while keeping encoder and transformers frozen (ablation)VQGAN decoder · Qualitative reconstructionvisual detailImproved
Muse-3B vs Stable Diffusion v1.4 on PartiPrompts with human ratersMuse-3B · PartiPrompts human evalprompt alignment preference: 70.6% vs 25.4%Improved
Best-of-16 reranking by CLIP score per promptMuse generations for human eval and examples · PartiPrompts / qualitativeselectionNo clear effect
Scaling the base model to 3B parameters with 48 layersMuse base model · MS-COCO zero-shotCLIP score: 0.32 vs Imagen 0.27Improved

Later papers that compared with this method

1 later accepted paper reports Muse-3B as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.

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

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy