Research map / 3D vision / Face and shape
PuppetGAN: Cross-Domain Image Manipulation by Demonstration
Ben Usman, Nick Dufour, Kate Saenko, Chris Bregler
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 |
|---|---|---|---|
| use PuppetGAN disentanglement losses instead of plain CycleGAN translation | PuppetGAN · MNIST digit size | correlation with synthetic attribute (r_attr): 0.28 -> 0.85 | Improved |
| use PuppetGAN disentanglement losses instead of plain CycleGAN translation | PuppetGAN · MNIST digit size | class accuracy preserved (Acc): 0.10 -> 0.73 | Improved |
| use PuppetGAN instead of DiDA | PuppetGAN · MNIST digit rotation | correlation with synthetic attribute (r_attr): 0.04 -> 0.40 | Improved |
| use PuppetGAN instead of MUNIT | PuppetGAN · MNIST digit rotation | correlation with synthetic attribute (r_attr): 0.00 -> 0.40 | Improved |
| use PuppetGAN instead of MUNIT | PuppetGAN · MNIST digit size | class accuracy preserved (Acc): 0.96 -> 0.73 | Hurt |
| use PuppetGAN instead of Cycle-VAE under domain shift | PuppetGAN · MNIST digit size | class accuracy preserved (Acc): 0.17 -> 0.73 | Improved |
| use PuppetGAN instead of Cycle-VAE | PuppetGAN · MNIST digit size | correlation with synthetic attribute (r_attr): 0.92 -> 0.85 | Hurt |
| increase discrepancy between synthetic and real attribute distributions (smaller synthetic digits) (ablation) | PuppetGAN · MNIST digit size | correlation with synthetic attribute (r_attr): 0.85 -> 0.28 | Hurt |
| increase discrepancy between synthetic and real attribute distributions (ablation) | PuppetGAN · MNIST digit rotation | class accuracy preserved (Acc): 0.97 -> 0.10 | Hurt |
Closest papers by meaning
- High Resolution Zero-Shot Domain Adaptation of Synthetically Rendered Face Images (ECCV 2020 · similarity 0.89)
- 3D Guided Fine-Grained Face Manipulation (CVPR 2019 · similarity 0.88)
- LeGO: Leveraging a Surface Deformation Network for Animatable Stylized Face Generation with One Example (CVPR 2024 · similarity 0.86)
- ControlFace: Harnessing Facial Parametric Control for Face Rigging (CVPR 2025 · similarity 0.85)
- General Articulated Objects Manipulation in Real Images via Part-Aware Diffusion Process (NeurIPS 2024 · similarity 0.85)
- RealPortrait: Realistic Portrait Animation with Diffusion Transformers (AAAI 2025 · similarity 0.85)
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.
