Research map / Image generation / Text-to-image diffusion
RiFeGAN: Rich Feature Generation for Text-to-Image Synthesis From Prior Knowledge
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 multiple original captions with SAEMs instead of concatenating captions without SAEMs (ablation) | RiFeGAN · CUB200 | Inception score: 4.90 ± 0.07 -> 5.23 ± 0.09 | Helped |
| Use multiple original captions with SAEMs instead of concatenating captions without SAEMs (ablation) | RiFeGAN · Oxford102 | Inception score: 4.23 ± 0.03 -> 4.53 ± 0.05 | Helped |
| Use multiple original captions with SAEMs instead of concatenating captions without SAEMs (ablation) | RiFeGAN · CUB200 | R-precision: 0.182 ± 0.012 -> 0.238± 0.015 | Helped |
| Use multiple original captions with SAEMs instead of concatenating captions without SAEMs (ablation) | RiFeGAN · Oxford102 | R-precision: 0.213 ± 0.015 -> 0.267± 0.016 | Helped |
| Use retrieved captions with SAEMs instead of retrieved captions without SAEMs (ablation) | RiFeGAN · CUB200 | Inception score: 4.79 ± 0.04 -> 4.85 ± 0.08 | Helped |
| Use retrieved captions with SAEMs instead of retrieved captions without SAEMs (ablation) | RiFeGAN · Oxford102 | Inception score: 4.09 ± 0.03 -> 4.23 ± 0.05 | Helped |
| Use retrieved captions with SAEMs instead of retrieved captions without SAEMs (ablation) | RiFeGAN · CUB200 | R-precision: 0.130 ± 0.012 -> 0.183± 0.013 | Helped |
| Use retrieved captions with SAEMs instead of retrieved captions without SAEMs (ablation) | RiFeGAN · Oxford102 | R-precision: 0.163± 0.014 -> 0.210 ± 0.014 | Helped |
| Increase the number of captions used for generation (ablation) | RiFeGAN · CUB200 and Oxford102 | Inception score | Helped |
| Repeat the corresponding caption to emphasize it in SAEMs (ablation) | RiFeGAN · CUB200 and Oxford102 | R-precision | Helped |
| Use two captions instead of one (ablation) | RiFeGAN · CUB200 and Oxford102 | R-precision | Hurt |
Closest papers by meaning
- LLM Blueprint: Enabling Text-to-Image Generation with Complex and Detailed Prompts (ICLR 2024 · similarity 0.86)
- CtrlSynth: Controllable Image Text Synthesis for Data-Efficient Multimodal Learning (ICML 2025 · similarity 0.85)
- IT3D: Improved Text-to-3D Generation with Explicit View Synthesis (AAAI 2024 · similarity 0.85)
- DreamText: High Fidelity Scene Text Synthesis (CVPR 2025 · similarity 0.85)
- Exploring the Deep Fusion of Large Language Models and Diffusion Transformers for Text-to-Image Synthesis (CVPR 2025 · similarity 0.85)
- Salient Concept-Aware Generative Data Augmentation (NeurIPS 2025 · similarity 0.85)
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.