Research map / Image generation / Text-to-image diffusion
Token Painter: Training-Free Text-Guided Image Inpainting via Mask Autoregressive Models
Longtao Jiang, Jie Huang, Mingfei Han, Lei Chen, Yongqiang Yu, Feng Zhao, Xiaojun Chang, Zhihui Li
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 changed | Where | Measured | Effect |
|---|---|---|---|
| add DEIF guidance token fusion to the T&B baseline (ablation) | Token Painter (NOVA) · BrushBench | PickScore x100: 19.47 -> 44.26 | Improved |
| add DEIF (ablation) | Token Painter (NOVA) · BrushBench | CLIP-S: 6.42 -> 14.42 | Improved |
| add guidance token enhancement (ADAE-G) (ablation) | Token Painter (NOVA) · BrushBench | PickScore x100: 44.26 -> 46.28 | Improved |
| add inpainting token enhancement (ADAE-I) (ablation) | Token Painter (NOVA) · BrushBench | PickScore x100: 46.28 -> 47.90 | Improved |
| add DEIF and ADAE (ablation) | Token Painter (NOVA) · BrushBench | PSNR: 26.26 -> 26.39 | No clear effect |
| use modified Gaussian instead of linear frequency fusion (ablation) | DEIF · BrushBench | PickScore x100: 44.84 -> 47.90 | Improved |
| use modified Gaussian instead of constant fusion (ablation) | DEIF · BrushBench | PickScore x100: 45.65 -> 47.90 | Improved |
| use modified Gaussian instead of quadratic fusion (ablation) | DEIF · BrushBench | PickScore x100: 46.42 -> 47.90 | Improved |
| use excessively large attention power exponents (ablation) | ADAE · qualitative | visual quality | Hurt |
| use T-only guidance instead of T&B (ablation) | MAR inpainting · qualitative | background harmony | Mixed |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 3 on most shared results.
- Meissonic — Meissonic: Revitalizing Masked Generative Transformers for Efficient High-Resolution Text-to-Image Synthesis (ICLR 2025) · better on 2 of 2 shared results
- BrushNet — BrushNet: A Plug-and-Play Image Inpainting Model with Decomposed Dual-Branch Diffusion (ECCV 2024) · better on 1 of 2 shared results
- HD-Painter — HD-Painter: High-Resolution and Prompt-Faithful Text-Guided Image Inpainting with Diffusion Models (ICLR 2025) · 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
- HD-Painter: High-Resolution and Prompt-Faithful Text-Guided Image Inpainting with Diffusion Models (ICLR 2025 · similarity 0.90)
- MTADiffusion: Mask Text Alignment Diffusion Model for Object Inpainting (CVPR 2025 · similarity 0.88)
- ATA: Adaptive Transformation Agent for Text-Guided Subject-Position Variable Background Inpainting (CVPR 2025 · similarity 0.87)
- One Stone with Two Birds: A Null-Text-Null Frequency-Aware Diffusion Models for Text-Guided Image Inpainting (NeurIPS 2025 · similarity 0.87)
- MaskINT: Video Editing via Interpolative Non-autoregressive Masked Transformers (CVPR 2024 · similarity 0.87)
- Towards Enhanced Image Inpainting: Mitigating Unwanted Object Insertion and Preserving Color Consistency (CVPR 2025 · similarity 0.87)
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.
