Research map / 3D vision / Face and shape
Synthesizing Coupled 3D Face Modalities by Trunk-Branch Generative Adversarial Networks
Baris Gecer, Alexandros Lattas, Stylianos Ploumpis, Jiankang Deng, Athanasios Papaioannou, Stylianos Moschoglou, Stefanos Zafeiriou
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 10K generated identities (Gen, 50 images each) to CASIA training data | ArcFace ResNet50 · CFP-FP | verification accuracy (%): 95.56% to 97.12% | Improved |
| add generated identities to CASIA training data | ArcFace ResNet50 · CFP-FP selected pair 1 | angle between pair (degrees): 84.06 -> 57.60 | Improved |
| add generated identities to CASIA training data | ArcFace ResNet50 · CFP-FP selected pair 4 | angle between pair (degrees): 88.06 -> 59.72 | Improved |
| add generated identities to CASIA training data | ArcFace ResNet50 · AgeDB-30 | verification accuracy | No clear effect |
| generate texture, shape and normals jointly with trunk-branch GAN instead of random shape or normals with fixed texture (ablation) | TBGAN · qualitative face realism | visual coherence | Improved |
| use TBGAN instead of PCA model for generation | 3D face generation · qualitative face realism | photorealism and detail | Improved |
| use TBGAN generator instead of PCA model | 3D face model · generator memory | memory (GB): 6GB -> 1.25GB | Improved |
| add generated normals to rendering of shape and texture (ablation) | TBGAN rendering · qualitative rendering | shading quality | Improved |
Closest papers by meaning
- 3DFaceGAN: Adversarial Nets for 3D Face Representation, Generation, and Translation (IJCV 2020 · similarity 0.92)
- Unsupervised High-Fidelity Facial Texture Generation and Reconstruction (ECCV 2022 · similarity 0.91)
- GANFIT: Generative Adversarial Network Fitting for High Fidelity 3D Face Reconstruction (CVPR 2019 · similarity 0.91)
- Photo-Realistic Facial Details Synthesis From Single Image (ICCV 2019 · similarity 0.90)
- Diffusion-driven GAN Inversion for Multi-Modal Face Image Generation (CVPR 2024 · similarity 0.89)
- LeGO: Leveraging a Surface Deformation Network for Animatable Stylized Face Generation with One Example (CVPR 2024 · 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.
