Research map / 3D vision / Human pose estimation
View Extrapolation of Human Body From a Single Image
Read the paperCheck 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 explicit depth prediction and geometric projection before flow refinement instead of directly learning view-to-view mapping | Human-body view synthesis · Synthetic human body test set | flow MSE: 34.04 -> 2.207 | Helped |
| Use explicit depth prediction and geometric projection before flow refinement instead of directly learning view-to-view mapping | Human-body view synthesis · Synthetic human body test set | image MSE: 131.6 -> 72.86 | Helped |
| Add flow-based mask prediction (ablation) | VSAP · Synthetic human body test set | image MSE: 131.6 -> 90.44 | Helped |
| Add flow-based mask prediction (ablation) | VSAP · Synthetic human body test set | mask IoU: 0.7903 -> 0.8907 | Helped |
| Add flow supervision (ablation) | VSAP+M. · Synthetic human body test set | flow MSE: 34.04 -> 13.01 | Helped |
| Add flow supervision (ablation) | VSAP+M. · Synthetic human body test set | image MSE: 90.44 -> 87.34 | Helped |
| Add flow supervision (ablation) | VSAP+M. · Synthetic human body test set | mask IoU: 0.8907 -> 0.8831 | Hurt |
| Use image loss plus flow loss instead of image loss alone (ablation) | Flow prediction network · Human-body view synthesis | prediction accuracy | Helped |
| Use a spatial transformer to predict a binary mask instead of an encoder-decoder softmax mask (ablation) | Mask prediction network · Human-body view synthesis | mask quality | Helped |
| Predict only residual mask regions using the transformed forward-flow mask (ablation) | Mask prediction network · Human-body view synthesis | mask quality | Helped |
| Use forward flow rather than source mask plus transforming vector as mask-network input (ablation) | Mask prediction network · Human-body view synthesis | mask prediction accuracy | No clear effect |
| Apply the proposed geometry-based method rather than VSAP | View synthesis · ShapeNet car | SSIM: 0.9061 -> 0.9058 | Hurt |
Closest papers by meaning
- Neural Body: Implicit Neural Representations With Structured Latent Codes for Novel View Synthesis of Dynamic Humans (CVPR 2021 · similarity 0.89)
- Human Shape From Silhouettes Using Generative HKS Descriptors and Cross-Modal Neural Networks (CVPR 2017 · similarity 0.89)
- Learning to Estimate 3D Human Pose and Shape From a Single Color Image (CVPR 2018 · similarity 0.88)
- Convolutional Mesh Regression for Single-Image Human Shape Reconstruction (CVPR 2019 · similarity 0.88)
- Neural Localizer Fields for Continuous 3D Human Pose and Shape Estimation (NeurIPS 2024 · similarity 0.87)
- Generalizable Human Gaussians from Single-View Image (ICLR 2025 · similarity 0.86)
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.