Research map / 3D vision / Depth and stereo
Efficient Multi-View Stereo by Iterative Dynamic Cost Volume
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 3 rather than 1 iteration at each stage (ablation) | Iterative dynamic cost volume method · DTU evaluation set | Overall (mm): 0.324 -> 0.317 | Helped |
| Add the initial depth prediction module instead of initializing from the median inverse depth (ablation) | GRU-based optimizer · DTU evaluation set | Overall (mm): 0.344 -> 0.317 | Helped |
| Add both depth and context features to the dynamic cost volume (ablation) | Iterative dynamic cost volume method · DTU evaluation set | Overall (mm): 0.347 -> 0.317 | Helped |
| Add depth features to the dynamic cost volume (ablation) | Iterative dynamic cost volume method · DTU evaluation set | Overall (mm): 0.340 -> 0.317 | Helped |
| Add context features to the dynamic cost volume (ablation) | Iterative dynamic cost volume method · DTU evaluation set | Overall (mm): 0.322 -> 0.317 | Helped |
| Use 4 rather than 1 iterations at each stage (ablation) | Iterative dynamic cost volume method · DTU evaluation set | Comp. (mm): 0.334 -> 0.311 | Helped |
| Use 4 rather than 1 iterations at each stage (ablation) | Iterative dynamic cost volume method · DTU evaluation set | Acc. (mm): 0.314 -> 0.323 | Hurt |
| Use 4 rather than 2 stages (ablation) | Iterative dynamic cost volume method · DTU evaluation set | Overall: 0.339 -> 0.314 | Helped |
| Use 4 rather than 2 stages (ablation) | Iterative dynamic cost volume method · Tanks & Temples | Mean F-score: 54.5 -> 58.8 | Helped |
| Use 4 rather than 2 stages (ablation) | Iterative dynamic cost volume method · DTU evaluation set | Time (s): 0.14 -> 0.31 | Hurt |
| Use 4 rather than 2 stages (ablation) | Iterative dynamic cost volume method · DTU evaluation set | Mem. (GB): 2.7 -> 4.6 | Hurt |
| Use 3 rather than 1 iterations at each stage (ablation) | Iterative dynamic cost volume method · DTU evaluation set | Time (s): 0.11 -> 0.19 | Hurt |
Closest papers by meaning
- Cascade Cost Volume for High-Resolution Multi-View Stereo and Stereo Matching (CVPR 2020 · similarity 0.91)
- IterMVS: Iterative Probability Estimation for Efficient Multi-View Stereo (CVPR 2022 · similarity 0.89)
- Cost Volume Pyramid Based Depth Inference for Multi-View Stereo (CVPR 2020 · similarity 0.89)
- Generalized Binary Search Network for Highly-Efficient Multi-View Stereo (CVPR 2022 · similarity 0.88)
- MVS2D: Efficient Multi-View Stereo via Attention-Driven 2D Convolutions (CVPR 2022 · similarity 0.87)
- MVSFormer++: Revealing the Devil in Transformer's Details for Multi-View Stereo (ICLR 2024 · similarity 0.84)
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.