Visual Attention Guided Learning with Incomplete Labels for Seismic Fault Interpretation

• Published in: IEEE transactions on geoscience and remote sensing Vol. 62; p. 1
• Main Authors: Mustafa, Ahmad, Rastegar, Reza, Brown, Tim, Nunes, Gregory, DeLilla, Daniel, AlRegib, Ghassan
• Format: Journal Article
• Language: English
• Published: New York IEEE 01-01-2024; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Annotations; Artificial neural networks; Attention; Cognitive ability; Data models; deep learning; fault interpretation; Fault lines; Faults; finetuning; Geological faults; Labels; Learning; Neural networks; Pixels; Seismic data; Surveys; Task analysis; Three-dimensional displays; Training; visual attention; Visual discrimination learning; Visual perception; Visualization
• ISSN: 0196-2892; 1558-0644
• DOI: 10.1109/TGRS.2024.3370037

Annotating geological faults on three dimensional seismic volumes is a laborious process. Typically, only a fraction of the actual faults are manually interpreted, leaving many others unlabeled. This is due to the way attention selectivity works to drive human perception. The human brain selectively focuses its attention to certain salient regions in a visual scene marked by prominent changes in color, contrast, and other low level signal cues. This bottom-up attention is further modulated by the individual's goals, expectations, and constraints with respect to the task at hand, also called top-down attention. The fault annotations created by seismic interpreters reflect this cognitive process comprising of both bottom-up and top-down attentional mechanisms. 3D convolutional neural networks pretrained on synthetic seismic data for fault mapping can be finetuned on select seismic lines extracted and labeled on a real seismic volume of interest. Traditional finetuning approaches treat all pixels on labeled sections as the absolute ground-truth. This leads to the network incorrectly learning to predict regions of missing fault labels as negatives. We propose an attention-guided training framework that models and incorporates human visual attention to (1) condition the process of sampling training data and (2) modulate the loss value for each pixel. Through quantitative and qualitative evaluation of results on a real seismic volume from North Western Australia, we demonstrate that the proposed approach is able to predict both the annotated as well as the unlabeled faults significantly better compared to baseline approaches.