Seismic demultiple with deep learning

Seismic data processing relies on multiples attenuation to improve inversion and interpretation. Radon-based algorithms are often used for multiples and primaries discrimination. Deep learning, based on convolutional neural networks (CNNs), has shown encouraging applications for demultiple that could mitigate Radon-based challenges. In this work, we investigate new strategies to train a CNN for multiples removal based on different loss functions. We propose combined primaries and multiples labels in the loss for training a CNN to predict primaries, multiples, or both simultaneously. Moreover, we investigate two distinctive training methods for all the strategies: UNet based on minimum absolute error (L1) training, and adversarial training (GAN-UNet). We test the trained models with the different strategies and methods on 400 synthetic data. We found that training to predict multiples, including the primaries labels as a constraint, is the most effective strategy for demultiple. Finally, we test the methods on a field dataset. UNet models trained with L1 report competitive results with Radon demultiple on field data, and better performance than GAN-UNet on synthetic and field data. As a result, it can potentially replace Radon-based demultiple in existing workflows.