Research map / Image generation / Editing and style transfer
CooGAN: A Memory-Efficient Framework for High-Resolution Facial Attribute Editing
Xuanhong Chen, Bingbing Ni, Naiyuan Liu, Ziang Liu, Yiliu Jiang, Loc Truong, Qi Tian
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 skip connections (0 to 4) (ablation) | U-Net style editor · CelebA | PSNR: 20.2 -> 34.8 | Improved |
| add skip connections (0 to 4) (ablation) | U-Net style editor · CelebA | average attribute editing accuracy: 89.5% -> 63.9% | Hurt |
| add the global module to the local module (ablation) | CooGAN · CelebA-HQ 768x768 | qualitative attribute editing | Improved |
| replace STU with LSTU (ablation) | LSTU-Net · skip connection (16 in-channel) | parameters: 1.63M -> 0.81M | Improved |
| replace STU with LSTU (ablation) | LSTU-Net · skip connection (16 in-channel) | FLOPS: 5.06G -> 2.02G | Improved |
| replace STU with LSTU (ablation) | LSTU-Net · skip connection (64 in-channel) | FLOPS: 79.50G -> 30.80G | Improved |
| replace STU with LSTU (ablation) | STGAN · CelebA 128x128 reconstruction | PSNR: 31.6 -> 32.1 | No clear effect |
| replace STU with LSTU (ablation) | STGAN · CelebA 128x128 reconstruction | SSIM: 0.934 -> 0.938 | No clear effect |
| use patch-based CooGAN instead of AttGAN-HR | CooGAN · CelebA-HQ 768x768 | GPU memory: 3692 MB -> 985 MB | Improved |
| use patch-based CooGAN instead of STGAN-HR | CooGAN · CelebA-HQ 768x768 | GPU memory: 6306 MB -> 985 MB | Improved |
| use patch-based CooGAN instead of STGAN-HR | CooGAN · CelebA-HQ 768x768 | inference time: 0.0502 s -> 0.0645 s | Hurt |
| use CooGAN instead of STGAN-HR | CooGAN · User study Male attribute | user preference: 51.3% -> 87.9% | 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.
- StarGAN — StarGAN: Unified Generative Adversarial Networks for Multi-Domain Image-to-Image Translation (CVPR 2018) · better on 2 of 2 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
- L2M-GAN: Learning To Manipulate Latent Space Semantics for Facial Attribute Editing (CVPR 2021 · similarity 0.89)
- Not Just Compete, but Collaborate: Local Image-to-Image Translation via Cooperative Mask Prediction (CVPR 2021 · similarity 0.88)
- Image-to-Image Translation With Disentangled Latent Vectors for Face Editing (TPAMI 2023 · similarity 0.88)
- STGAN: A Unified Selective Transfer Network for Arbitrary Image Attribute Editing (CVPR 2019 · similarity 0.87)
- A Latent Transformer for Disentangled Face Editing in Images and Videos (ICCV 2021 · similarity 0.87)
- SDGAN: Disentangling Semantic Manipulation for Facial Attribute Editing (AAAI 2024 · 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.
