Research map / 3D vision / Face and shape
OSTeC: One-Shot Texture Completion
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 one-shot optimization with a 2D face generator and progressive UV stitching instead of CE | OSTeC · MultiPIE UVDB ±90° | PSNR: 19.63 -> 20.44 | Helped |
| Use one-shot optimization with a 2D face generator and progressive UV stitching instead of UV-GAN | OSTeC · MultiPIE UVDB ±90° | SSIM: 0.7250 -> 0.7462 | Helped |
| Frontalize faces before recognition | ResNet-18 trained on MS1M with ArcFace · Multi-PIE ±90° | Rank-1 recognition rate: 92.06 -> 95.24 | Helped |
| Add profile-to-frontal synthesis | CASIA-R18-ArcFace · CFP Frontal-Profile | Verification accuracy: 93.69 -> 94.87 | Helped |
| Add profile-to-frontal synthesis | MS1M-R18-ArcFace · CFP Frontal-Profile | Verification accuracy: 96.14 -> 97.06 | Helped |
| Add frontal-to-profile synthesis | CASIA-R18-ArcFace · CFP Frontal-Profile | Verification accuracy: 93.69 -> 95.68 | Helped |
| Add frontal-to-profile synthesis | MS1M-R18-ArcFace · CFP Frontal-Profile | Verification accuracy: 96.14 -> 97.43 | Helped |
| Use view interpolation and set-centre verification | CASIA-R18-ArcFace · CFP Frontal-Profile | Verification accuracy: 93.69 -> 95.92 | Helped |
| Use view interpolation and set-centre verification | MS1M-R18-ArcFace · CFP Frontal-Profile | Verification accuracy: 96.14 -> 97.85 | Helped |
| Add photometric loss to encoder initialization (ablation) | OSTeC · Texture completion optimization | identity/visual reconstruction quality | Helped |
| Add landmark loss (ablation) | OSTeC · Texture completion optimization | landmark alignment | Helped |
| Add perceptual loss (ablation) | OSTeC · Texture completion optimization | mid-level feature similarity | Helped |
Closest papers by meaning
- GANFIT: Generative Adversarial Network Fitting for High Fidelity 3D Face Reconstruction (CVPR 2019 · similarity 0.90)
- Towards High-Fidelity 3D Face Reconstruction From In-the-Wild Images Using Graph Convolutional Networks (CVPR 2020 · similarity 0.89)
- Towards High-Fidelity Face Self-Occlusion Recovery via Multi-View Residual-Based GAN Inversion (AAAI 2022 · similarity 0.88)
- Next3D: Generative Neural Texture Rasterization for 3D-Aware Head Avatars (CVPR 2023 · similarity 0.88)
- FreeUV: Ground-Truth-Free Realistic Facial UV Texture Recovery via Cross-Assembly Inference Strategy (CVPR 2025 · similarity 0.86)
- UV-IDM: Identity-Conditioned Latent Diffusion Model for Face UV-Texture Generation (CVPR 2024 · 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.