A Principled Benchmark for Seismic Data Segmentation
In recent years, several deep learning techniques have been applied to the problem of seismic facies segmentation. However, there is a lack of an authoritative protocol for evaluating such models, so that the comparison between results becomes compromised. This paper proposes a principled benchmark for lithofacies segmentation based on the public seismic volumes from the F3 Netherlands, Penobscot and Parihaka datasets, along with standard metrics for performance assessment. The utility of the benchmark is illustrated by the evaluation of the U-Net DeconvNet and SegNet encoder-decoder architectures. The goal is to offer a framework which will enable researchers to compare different methods and to develop more effective strategies for the segmentation problem.
Authors
- Luiz Antônio de Lima (ORCID: https://orcid.org/0000-0003-4228-2387)
- Pedro O. S. Vaz de Melo (ORCID: https://orcid.org/0000-0002-9749-0151)
- Alexei Manso Corrêa Machado (ORCID: https://orcid.org/0000-0001-8077-3377)
- Gabriela Nunes Sales Barreto (ORCID: https://orcid.org/0000-0002-1906-7737)
- Cláudia Mara de Melo Tavares (ORCID: https://orcid.org/0000-0002-8416-6272)
- William Robson Schwartz (ORCID: https://orcid.org/0000-0003-1449-8834)
- Thales H. Silva
- Gabriel L. Canguçu
- Leonardo M. S. Jorge
- Walace S. Caldas
Institutions
- Universidade Federal de Minas Gerais (BR)
- Pontifícia Universidade Católica de Minas Gerais (BR)
- Petrobras (Brazil) (BR)
Publication Details
- Journal
- IEEE Transactions on Geoscience and Remote Sensing
- Published
- 2025-01-01
- DOI
- https://doi.org/10.1109/tgrs.2025.3575336
- Citations
- 1
- Primary Topic
- Seismic Imaging and Inversion Techniques
- Type
- article
- Field-Weighted Citation Impact
- 1.87