Adaptive Learning for Multi-view Stereo Reconstruction
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2024
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866913305184960512 |
|---|---|
| author | Min, Qinglu Zhao, Jie Zhang, Zhihao Min, Chen |
| author_facet | Min, Qinglu Zhao, Jie Zhang, Zhihao Min, Chen |
| contents | Deep learning has recently demonstrated its excellent performance on the task of multi-view stereo (MVS). However, loss functions applied for deep MVS are rarely studied. In this paper, we first analyze existing loss functions' properties for deep depth based MVS approaches. Regression based loss leads to inaccurate continuous results by computing mathematical expectation, while classification based loss outputs discretized depth values. To this end, we then propose a novel loss function, named adaptive Wasserstein loss, which is able to narrow down the difference between the true and predicted probability distributions of depth. Besides, a simple but effective offset module is introduced to better achieve sub-pixel prediction accuracy. Extensive experiments on different benchmarks, including DTU, Tanks and Temples and BlendedMVS, show that the proposed method with the adaptive Wasserstein loss and the offset module achieves state-of-the-art performance. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2404_05181 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Adaptive Learning for Multi-view Stereo Reconstruction Min, Qinglu Zhao, Jie Zhang, Zhihao Min, Chen Computer Vision and Pattern Recognition Deep learning has recently demonstrated its excellent performance on the task of multi-view stereo (MVS). However, loss functions applied for deep MVS are rarely studied. In this paper, we first analyze existing loss functions' properties for deep depth based MVS approaches. Regression based loss leads to inaccurate continuous results by computing mathematical expectation, while classification based loss outputs discretized depth values. To this end, we then propose a novel loss function, named adaptive Wasserstein loss, which is able to narrow down the difference between the true and predicted probability distributions of depth. Besides, a simple but effective offset module is introduced to better achieve sub-pixel prediction accuracy. Extensive experiments on different benchmarks, including DTU, Tanks and Temples and BlendedMVS, show that the proposed method with the adaptive Wasserstein loss and the offset module achieves state-of-the-art performance. |
| title | Adaptive Learning for Multi-view Stereo Reconstruction |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2404.05181 |