Research map / 3D vision / Depth and stereo
CL-MVSNet: Unsupervised Multi-View Stereo with Dual-Level Contrastive Learning
Kaiqiang Xiong, Rui Peng, Zhe Zhang, Tianxing Feng, Jianbo Jiao, Feng Gao, Ronggang Wang
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 image-level and scene-level contrastive consistency branches and L0.5 loss to photometric plus augmentation baseline (ablation) | unsupervised MVS · DTU evaluation set | Overall distance (mm): 0.378 -> 0.329 | Improved |
| replace vanilla photometric loss with L0.5 photometric consistency loss (ablation) | CL-MVSNet · DTU evaluation set | Overall distance (mm): 0.339 -> 0.329 | Improved |
| use CL-MVSNet rather than prior end-to-end unsupervised method RC-MVSNet | unsupervised MVS · DTU evaluation set | Overall distance (mm): 0.345 -> 0.329 | Improved |
| use CL-MVSNet instead of supervised CasMVSNet | DTU · DTU evaluation set | Overall distance (mm): 0.355 -> 0.329 | Improved |
| use CL-MVSNet instead of RC-MVSNet on Tanks&Temples without finetuning | unsupervised MVS · Tanks&Temples Intermediate | mean F-score: 55.04 -> 59.39 | Improved |
| compare end-to-end CL-MVSNet to multi-stage self-supervised KD-MVS | Tanks&Temples · Tanks&Temples Intermediate | mean F-score: 64.14 -> 59.39 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- RC-MVSNet — RC-MVSNet: Unsupervised Multi-View Stereo with Neural Rendering (ECCV 2022) · better on 4 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
- RC-MVSNet: Unsupervised Multi-View Stereo with Neural Rendering (ECCV 2022 · similarity 0.92)
- MVS2D: Efficient Multi-View Stereo via Attention-Driven 2D Convolutions (CVPR 2022 · similarity 0.91)
- "Smoothness, Synthesis, and Sampling: Re-thinking Unsupervised Multi-View Stereo with DIV Loss" (ECCV 2024 · similarity 0.91)
- MVSFormer++: Revealing the Devil in Transformer's Details for Multi-View Stereo (ICLR 2024 · similarity 0.90)
- Hierarchical Object-Aware Dual-Level Contrastive Learning for Domain Generalized Stereo Matching (NeurIPS 2024 · similarity 0.89)
- MVSAnywhere: Zero-Shot Multi-View Stereo (CVPR 2025 · similarity 0.89)
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.
