Research map / 3D vision / 3D generation
Rotationally-Temporally Consistent Novel View Synthesis of Human Performance Video
Youngjoong Kwon, Stefano Petrangeli, Dahun Kim, Haoliang Wang, Eunbyung Park, Viswanathan Swaminathan, Henry Fuchs
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 rotational consistency supervision (ablation) | Base siamese network · MVHA ablation (seen model, unseen pose) | L1: .0825 -> .0736 | Improved |
| add rotational consistency supervision (ablation) | Base siamese network · MVHA ablation (seen model, unseen pose) | Rotational loss: .0433 -> .0390 | Improved |
| add rotational consistency supervision (ablation) | Base siamese network · MVHA ablation (seen model, unseen pose) | Temporal loss: .0280 -> .0163 | Improved |
| add rotational consistency supervision (ablation) | Base siamese network · MVHA ablation (seen model, unseen pose) | SSIM: .7364 -> .7541 | Improved |
| add temporal consistency supervision (ablation) | Base siamese network · MVHA ablation (seen model, unseen pose) | Temporal loss: .0280 -> .0149 | Improved |
| add temporal consistency supervision (ablation) | Base siamese network · MVHA ablation (seen model, unseen pose) | L1: .0825 -> .0743 | Improved |
| add temporal consistency supervision (ablation) | Base siamese network · MVHA ablation (seen model, unseen pose) | Rotational loss: .0433 -> .0414 | Improved |
| add both rotational and temporal consistency (ablation) | Base siamese network · MVHA ablation (seen model, unseen pose) | L1: .0825 -> .0722 | Improved |
| add both rotational and temporal consistency (ablation) | Base siamese network · MVHA ablation (seen model, unseen pose) | Temporal loss: .0280 -> .0146 | Improved |
| add both rotational and temporal consistency (ablation) | Base siamese network · MVHA ablation (seen model, unseen pose) | SSIM: .7364 -> .7596 | Improved |
| add rotational and temporal consistency on top of TC alone (ablation) | Base siamese network · MVHA ablation (seen model, unseen pose) | L1: .0743 -> .0722 | Improved |
| use proposed siamese network with gating instead of TBN | Ours · MVHA (unseen model, unseen pose) | L1: .1130 -> .0863 | Improved |
Closest papers by meaning
- Diffuman4D: 4D Consistent Human View Synthesis from Sparse-View Videos with Spatio-Temporal Diffusion Models (ICCV 2025 · similarity 0.89)
- Scaling4D: Pushing the Frontier of Video Novel View Synthesis through Large-Scale Monocular Videos (CVPR 2026 · similarity 0.88)
- SimVS: Simulating World Inconsistencies for Robust View Synthesis (CVPR 2025 · similarity 0.87)
- Evaluating Multiview Object Consistency in Humans and Image Models (NeurIPS 2024 · similarity 0.87)
- ViVid-1-to-3: Novel View Synthesis with Video Diffusion Models (CVPR 2024 · similarity 0.86)
- MVSplat360: Feed-Forward 360 Scene Synthesis from Sparse Views (NeurIPS 2024 · 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.
