Research map / Image generation / Editing and style transfer
TediGAN: Text-Guided Diverse Face Image Generation and Manipulation
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 TediGAN instead of the strongest reported baseline, ControlGAN | Text-to-image generation · Multi-modal CelebA-HQ | FID: 116.32 -> 106.37 | Helped |
| Use TediGAN instead of AttnGAN | Text-to-image generation · Multi-modal CelebA-HQ | LPIPS: 0.512 -> 0.456 | Helped |
| Use TediGAN instead of the strongest reported baseline, DFGAN | Text-to-image generation · Multi-modal CelebA-HQ | accuracy (%): 22.8 -> 25.3 | Helped |
| Use TediGAN instead of the strongest reported baseline, DFGAN | Text-to-image generation · Multi-modal CelebA-HQ | realism (%): 25.5 -> 31.7 | Helped |
| Use TediGAN instead of ManiGAN | Text-guided image manipulation · CelebA | FID: 117.89 -> 107.25 | Helped |
| Use TediGAN instead of ManiGAN | Text-guided image manipulation · CelebA | accuracy (%): 40.9 -> 59.1 | Helped |
| Use TediGAN instead of ManiGAN | Text-guided image manipulation · CelebA | realism (%): 36.2 -> 63.8 | Helped |
| Use TediGAN instead of ManiGAN | Text-guided image manipulation · Non-CelebA | FID: 143.39 -> 135.47 | Helped |
| Use TediGAN instead of ManiGAN | Text-guided image manipulation · Non-CelebA | accuracy (%): 12.8 -> 87.2 | Helped |
| Use TediGAN instead of ManiGAN | Text-guided image manipulation · Non-CelebA | realism (%): 21.7 -> 78.3 | Helped |
| Add instance-level optimization to the image-encoder inversion (ablation) | TediGAN image manipulation · Image manipulation; qualitative identity preservation | identity preservation | Helped |
| Use the visual-linguistic similarity and pairwise ranking strategy for text-image alignment (ablation) | TediGAN text-guided synthesis · Near-miss text-image cases | attribute alignment | Mixed |
Closest papers by meaning
- When StyleGAN Meets Stable Diffusion: a W+ Adapter for Personalized Image Generation (CVPR 2024 · similarity 0.86)
- Fast Text-to-3D-Aware Face Generation and Manipulation via Direct Cross-modal Mapping and Geometric Regularization (ICML 2024 · similarity 0.86)
- Diffusion-driven GAN Inversion for Multi-Modal Face Image Generation (CVPR 2024 · similarity 0.85)
- Text-conditional Attribute Alignment across Latent Spaces for 3D Controllable Face Image Synthesis (CVPR 2024 · similarity 0.85)
- DreamSalon: A Staged Diffusion Framework for Preserving Identity-Context in Editable Face Generation (CVPR 2024 · similarity 0.84)
- Learning to Manipulate Artistic Images (AAAI 2024 · similarity 0.84)
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.