Attention-Based 3D Seismic Fault Segmentation Training by a Few 2D Slice Labels

TL;DR

This paper introduces a novel method for 3D seismic fault segmentation that leverages limited 2D slice labels and an attention module to effectively learn from noisy, complex seismic data, reducing labeling effort.

Contribution

It proposes lambda-BCE and lambda-smooth L1loss functions, an attention module for noise suppression, and demonstrates effective 3D fault segmentation from minimal 2D annotations.

Findings

Effective learning from only 3.3% of 3D labels

Attention module significantly suppresses seismic noise

Achieves comparable performance to full 3D labeling

Abstract

Detection faults in seismic data is a crucial step for seismic structural interpretation, reservoir characterization and well placement. Some recent works regard it as an image segmentation task. The task of image segmentation requires huge labels, especially 3D seismic data, which has a complex structure and lots of noise. Therefore, its annotation requires expert experience and a huge workload. In this study, we present lambda-BCE and lambda-smooth L1loss to effectively train 3D-CNN by some slices from 3D seismic data, so that the model can learn the segmentation of 3D seismic data from a few 2D slices. In order to fully extract information from limited data and suppress seismic noise, we propose an attention module that can be used for active supervision training and embedded in the network. The attention heatmap label is generated by the original label, and letting it supervise the attention module using the lambda-smooth L1loss. The experiment demonstrates the effectiveness of our loss function, the method can extract 3D seismic features from a few 2D slice labels. And it also shows the advanced performance of the attention module, which can significantly suppress the noise in the seismic data while increasing the model's sensitivity to the foreground. Finally, on the public test set, we only use the 2D slice labels training that accounts for 3.3% of the 3D volume label, and achieve similar performance to the 3D volume label training.