Automatic classification of geologic units in seismic images using partially interpreted examples
This work addresses the efficiency challenge for seismic interpreters by making neural network-based semantic segmentation more practical, though it is incremental as it builds on existing image-to-image mapping networks.
The paper tackles the problem of time-consuming manual labeling for seismic image interpretation by introducing a partial loss-function and labeling strategies that enable neural networks to learn from partially interpreted images, achieving high-quality predictions with only a small number of annotated pixels per image.
Geologic interpretation of large seismic stacked or migrated seismic images can be a time-consuming task for seismic interpreters. Neural network based semantic segmentation provides fast and automatic interpretations, provided a sufficient number of example interpretations are available. Networks that map from image-to-image emerged recently as powerful tools for automatic segmentation, but standard implementations require fully interpreted examples. Generating training labels for large images manually is time consuming. We introduce a partial loss-function and labeling strategies such that networks can learn from partially interpreted seismic images. This strategy requires only a small number of annotated pixels per seismic image. Tests on seismic images and interpretation information from the Sea of Ireland show that we obtain high-quality predicted interpretations from a small number of large seismic images. The combination of a partial-loss function, a multi-resolution network that explicitly takes small and large-scale geological features into account, and new labeling strategies make neural networks a more practical tool for automatic seismic interpretation.