Research map / Image generation / Text-to-image diffusion
TF-TI2I: Training-Free Text-and-Image-to-Image Generation via Multi-Modal Implicit-Context Learning In Text-to-Image Models
Teng-Fang Hsiao, Bo-Kai Ruan, Yi-Lun Wu, Tzu-Ling Lin, Hong-Han Shuai
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 |
|---|---|---|---|
| share contextual (text) tokens instead of concatenating reference vision tokens | TF-TI2I · attention compute | compute and memory: (2R-1) times vs ((R-1)+(n_P/n_I)R) times | Improved |
| add Reference Contextual Masking (RCM) to contextual token sharing (ablation) | TF-TI2I · qualitative multi-reference example | reference fidelity | Improved |
| add Winner-Takes-All (WTA) module (ablation) | TF-TI2I · qualitative multi-reference example | detail and color richness | Improved |
| use TF-TI2I instead of OmniGen on single-reference FG-TI2I | TF-TI2I · FG-TI2I OBJ(text) x ACT(img) | Image Reward: 0.85 -> 1.30 | Improved |
| TF-TI2I vs MasaCtrl on reference alignment | TF-TI2I · FG-TI2I OBJ(img) x TEX(text) | DINO similarity: 96.4 -> 87.5 | Hurt |
| TF-TI2I vs Kosmos-G on DreamBench reference alignment | TF-TI2I · DreamBench | CLIP-Image: 84.7 -> 79.1 | Hurt |
| TF-TI2I vs best editing baseline on Wild-TI2I | TF-TI2I · Wild-TI2I | CLIP-Directional Score: 0.210 -> 0.115 | Hurt |
| TF-TI2I vs best editing baseline on Wild-TI2I | TF-TI2I · Wild-TI2I | CLIP-Text: 29.3 -> 30.4 | 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.
- OmniGen — OmniGen: Unified Image Generation (CVPR 2025) · better on 6 of 7 shared results
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
- Learning Visual Generative Priors without Text (CVPR 2025 · similarity 0.91)
- Text-Printed Image: Bridging the Image-Text Modality Gap for Text-centric Training of Large Vision-Language Models (CVPR 2026 · similarity 0.89)
- MARS: Mixture of Auto-Regressive Models for Fine-grained Text-to-image Synthesis (AAAI 2025 · similarity 0.89)
- TokenFlow: Unified Image Tokenizer for Multimodal Understanding and Generation (CVPR 2025 · similarity 0.89)
- T2I-Adapter: Learning Adapters to Dig Out More Controllable Ability for Text-to-Image Diffusion Models (AAAI 2024 · similarity 0.88)
- OneIG-Bench: Omni-dimensional Nuanced Evaluation for Image Generation (NeurIPS 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.
