Research map / 3D vision / Face and shape
Deep3D: Fully Automatic 2D-to-3D Video Conversion with Deep Convolutional Neural Networks
Junyuan Xie, Ross Girshick, Ali Farhadi
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 lower-level feature side branches (multi-level upsampled features) (ablation) | Deep3D · 3D movie test set | MAE: 8.24 -> 6.87 | Improved |
| train end-to-end directly on stereo pairs instead of on block-matching disparity maps (ablation) | Deep3D · 3D movie test set | MAE: 7.29 -> 6.87 | Improved |
| add the disparity representation with differentiable selection layer instead of direct regression (ablation) | Deep3D · 3D movie test set | MAE: 7.01 -> 6.87 | Improved |
| use 5 consecutive RGB frames as input instead of a single frame (ablation) | Deep3D · 3D movie test set | MAE: 6.87 -> 6.81 | Improved |
| add 5 optical flow frames to a single RGB frame input (ablation) | Deep3D · 3D movie test set | MAE: 6.87 -> 6.86 | No clear effect |
| use Deep3D instead of Eigen et al. depth plus DIBR (fixed B and f) | 2D-to-3D conversion · 3D movie test set | MAE: 7.75 -> 6.87 | Improved |
| use Deep3D instead of Eigen et al. depth plus DIBR (oracle focus plane) | 2D-to-3D conversion · 3D movie test set (oracle) | MAE: 6.31 -> 5.47 | Improved |
| use Deep3D instead of Eigen et al. + Oracle in pairwise human judgments | 2D-to-3D conversion · Human study | preference rate (%): 66.08% preferred Deep3D over Eigen et al. + Oracle (33.92% the reverse) | Improved |
| use Deep3D instead of Global Disparity in pairwise human judgments | 2D-to-3D conversion · Human study | preference rate (%): 74.58% preferred Deep3D over Global Disparity (25.42% the reverse) | Improved |
Later papers that compared with this method
8 later accepted papers report Deep3D as a baseline in their results tables; 7 of them report a better number on most of the results they share with it.
- XPaintNet: An eXtreme Lightweight Framework for Stereoscopic Conversion without Inpainting Network (CVPR 2026 · better on 3 of 3 shared results)
- Monocular Avatar Reconstruction via Cascaded Diffusion Priors and UV-Space Differentiable Shading (ECCV 2026 · better on 1 of 1 shared result)
- 3D Multi-frame Fusion for Video Stabilization (CVPR 2024 · better on 0 of 4 shared results)
- A Hierarchical Representation Network for Accurate and Detailed Face Reconstruction From In-the-Wild Images (CVPR 2023 · better on 2 of 2 shared results)
- HiFace: High-Fidelity 3D Face Reconstruction by Learning Static and Dynamic Details (ICCV 2023 · better on 2 of 2 shared results)
- FFHQ-UV: Normalized Facial UV-Texture Dataset for 3D Face Reconstruction (CVPR 2023 · better on 1 of 1 shared result)
- Structure-Preserving Stereoscopic View Synthesis With Multi-Scale Adversarial Correlation Matching (CVPR 2019 · better on 6 of 6 shared results)
- A Lightweight Neural Network for Monocular View Generation With Occlusion Handling (TPAMI 2019 · better on 3 of 4 shared results)
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Elastic3D: Controllable Stereo Video Conversion with Guided Latent Decoding (CVPR 2026 · similarity 0.86)
- DeepMVS: Learning Multi-View Stereopsis (CVPR 2018 · similarity 0.86)
- DeepVideoMVS: Multi-View Stereo on Video With Recurrent Spatio-Temporal Fusion (CVPR 2021 · similarity 0.86)
- Learning-Based Multi-View Stereo: A Survey (TPAMI 2026 · similarity 0.86)
- DL3DV-10K: A Large-Scale Scene Dataset for Deep Learning-based 3D Vision (CVPR 2024 · similarity 0.85)
- SpatialDreamer: Self-supervised Stereo Video Synthesis from Monocular Input (CVPR 2025 · similarity 0.83)
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.
