Research map / Image generation / Editing and style transfer
Image-to-Image Translation With Disentangled Latent Vectors for Face Editing
Yusuf Dalva, Hamza Pehlivan, Öykü Irmak Hatipoğlu, Cansu Moran, Ayşegül Dündar
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 orthogonality loss (ablation) | VecGAN · CelebA-HQ Setting I (bangs addition, 128x128), latent-guided | FID: 21.98 -> 20.17 | Improved |
| add sparsity loss (ablation) | VecGAN · CelebA-HQ Setting I (bangs addition, 128x128), latent-guided | FID: 24.07 -> 20.17 | Improved |
| add sparsity loss (ablation) | VecGAN · CelebA-HQ Setting I (bangs addition, 128x128), reference-guided | FID: 22.43 -> 20.72 | Improved |
| add disentanglement loss (ablation) | VecGAN · CelebA-HQ Setting I (bangs addition, 128x128), reference-guided | FID: 20.72 -> 20.57 | Improved |
| add disentanglement loss (ablation) | VecGAN · CelebA-HQ Setting I (bangs addition, 128x128), latent-guided | FID: 20.17 -> 20.23 | No clear effect |
| add single-layer attention skip connection (ablation) | Disent. · CelebA-HQ Setting I (bangs addition, 128x128), reference-guided | FID: 20.57 -> 20.08 | Improved |
| replace single-layer attention skip with UNet attention skip (ablation) | Disent. + Attn. Skip · CelebA-HQ Setting I (bangs addition, 128x128), reference-guided | FID: 20.08 -> 19.87 | Improved |
| feed decoded features to UNet attention skip (ablation) | Disent. + Attn. UNet Skip · CelebA-HQ Setting I (bangs addition, 128x128), latent-guided | FID: 19.98 -> 19.65 | Improved |
| upgrade VecGAN to VecGAN++ (disentanglement loss and attention skips) | VecGAN · CelebA-HQ Setting I (bangs addition, 128x128), latent-guided | FID: 20.17 -> 19.65 | Improved |
| upgrade VecGAN to VecGAN++ | VecGAN · CelebA-HQ Setting II (256x256), bangs addition | FID: 36.47 -> 27.09 | Improved |
| upgrade VecGAN to VecGAN++ | VecGAN · CelebA-HQ Setting II (256x256), gender female | FID: 63.48 -> 30.24 | Improved |
| upgrade VecGAN to VecGAN++ | VecGAN · CelebA-HQ Setting II (256x256), blonde hair (+) | FID: 68.89 -> 37.18 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 on most shared results.
- StyleRes — StyleRes: Transforming the Residuals for Real Image Editing With StyleGAN (CVPR 2023) · better on 5 of 5 shared results
- StarGANv2 — StarGAN v2: Diverse Image Synthesis for Multiple Domains (CVPR 2020) · better on 2 of 2 shared results
- Elegant — ELEGANT: Exchanging Latent Encodings with GAN for Transferring Multiple Face Attributes (ECCV 2018) · better on 1 of 1 shared result
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 0.94)
- Interpreting the Latent Space of GANs for Semantic Face Editing (CVPR 2020 · similarity 0.93)
- Latent Transformations via NeuralODEs for GAN-Based Image Editing (ICCV 2021 · similarity 0.93)
- A Latent Transformer for Disentangled Face Editing in Images and Videos (ICCV 2021 · similarity 0.92)
- FEditNet++: Few-Shot Editing of Latent Semantics in GAN Spaces With Correlated Attribute Disentanglement (TPAMI 2024 · similarity 0.91)
- SDGAN: Disentangling Semantic Manipulation for Facial Attribute Editing (AAAI 2024 · similarity 0.91)
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.
