Research map / Image generation / Text-to-image diffusion
MirrorGAN: Learning Text-To-Image Generation by Redescription
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 MirrorGAN instead of AttnGAN | text-to-image generation · CUB | Inception Score: 4.36 -> 4.56 | Helped |
| use MirrorGAN instead of AttnGAN | text-to-image generation · COCO | Inception Score: 25.89 -> 26.47 | Helped |
| use MirrorGAN instead of AttnGAN | text-to-image generation · CUB | R-precision top-1: 53.31 -> 57.67 | Helped |
| use MirrorGAN instead of AttnGAN | text-to-image generation · COCO | R-precision top-1: 72.13 -> 74.52 | Helped |
| add STREAM reconstruction supervision (ablation) | MirrorGAN without global attention · CUB | Inception Score: 3.91 -> 4.47 | Helped |
| add STREAM reconstruction supervision (ablation) | MirrorGAN without global attention · COCO | Inception Score: 19.01 -> 25.99 | Helped |
| add global attention (ablation) | MirrorGAN with STREAM and local word attention · CUB | Inception Score: 4.47 -> 4.54 | Helped |
| add global attention (ablation) | MirrorGAN with STREAM and local word attention · COCO | Inception Score: 25.99 -> 26.47 | Helped |
| increase reconstruction-loss weight λ from 5 to 20 (ablation) | MirrorGAN · CUB | Inception Score: 4.01 -> 4.54 | Helped |
| increase reconstruction-loss weight λ from 5 to 20 (ablation) | MirrorGAN · COCO | R-precision top-1: 52.55% -> 74.52% | Helped |
| remove STREAM and global attention (ablation) | MirrorGAN · CUB | R-precision top-1: 57.67 -> 39.09 | Hurt |
| remove global attention (ablation) | MirrorGAN with STREAM · COCO | R-precision top-1: 74.52 -> 73.28 | Hurt |
Closest papers by meaning
- Learn, Imagine and Create: Text-to-Image Generation from Prior Knowledge (NeurIPS 2019 · similarity 0.89)
- Adversarial Text to Continuous Image Generation (CVPR 2024 · similarity 0.87)
- Leveraging Unpaired Data for Vision-Language Generative Models via Cycle Consistency (ICLR 2024 · similarity 0.87)
- Elucidating the design space of language models for image generation (ICML 2025 · similarity 0.86)
- LLM Blueprint: Enabling Text-to-Image Generation with Complex and Detailed Prompts (ICLR 2024 · similarity 0.86)
- Discriminative Probing and Tuning for Text-to-Image Generation (CVPR 2024 · similarity 0.86)
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.