Research map / Image generation / Editing and style transfer
Talk-to-Edit: Fine-Grained 2D and 3D Facial Editing via Dialog
Yuming Jiang, Ziqi Huang, Tianxing Wu, Xingang Pan, Chen Change Loy, Ziwei Liu
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 |
|---|---|---|---|
| replace straight-line InterfaceGAN direction with location-specific semantic field (curved trajectory) (ablation) | Talk-to-Edit · Bangs | identity preservation distance: 0.7621 -> 0.6047 | Improved |
| use location-specific semantic field instead of identity-agnostic Multiclass SVM directions (ablation) | Talk-to-Edit · Smiling | attribute preservation cross-entropy: 0.8676 -> 0.5028 | Improved |
| use semantic field instead of fixed per-identity direction (Enjoy Your Editing) (ablation) | Talk-to-Edit · Bangs | identity preservation distance: 0.6693 -> 0.6047 | Improved |
| edit in W+ space instead of Z space (ablation) | Talk-to-Edit · Smiling | identity preservation distance: 0.6434 -> 0.4580 | Improved |
| edit in W+ space instead of Z space (ablation) | Talk-to-Edit · Eyeglasses | identity preservation distance: 0.6229 -> 0.6670 | Hurt |
| edit in W+ space instead of Z space (ablation) | Talk-to-Edit · Young | attribute preservation cross-entropy: 0.4814 -> 0.2731 | Improved |
| add fine-grained labels instead of binary labels to supervise semantic field | Talk-to-Edit · facial editing | edit smoothness | Improved |
| add language dialog system on top of semantic field | Talk-to-Edit · user study | user preference: around 80% | Improved |
| add location-specific property (varying direction per step) (ablation) | Talk-to-Edit · eyeglasses on 1024x1024 (100 latent codes) | cosine similarity with initial direction | 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.
- InterfaceGAN — InterFaceGAN: Interpreting the Disentangled Face Representation Learned by GANs (TPAMI 2020) · better on 5 of 5 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
- Talk-To-Edit: Fine-Grained Facial Editing via Dialog (ICCV 2021 · similarity 0.99)
- SDGAN: Disentangling Semantic Manipulation for Facial Attribute Editing (AAAI 2024 · similarity 0.89)
- L2M-GAN: Learning To Manipulate Latent Space Semantics for Facial Attribute Editing (CVPR 2021 · similarity 0.88)
- FineEdit: Fine-Grained Image Edit with Bounding Box Guidance (ECCV 2026 · similarity 0.88)
- Editing in Style: Uncovering the Local Semantics of GANs (CVPR 2020 · similarity 0.88)
- Image-to-Image Translation With Disentangled Latent Vectors for Face Editing (TPAMI 2023 · similarity 0.88)
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.
