CORTEXA
← Browse
openalexZenodo (CERN European Organization for Nuclear Research)2026-07-24Cited by 0

Fast Prestack Wavefield Separation via 3D U-Net Knowledge Distillation

Alexander Shcherbina

We present a hybrid approach for automated prestack seismic data processing that combines expert-configured High-Resolution Wavefield Separation (HRWS) algorithms with 3D U-Net neural networks through knowledge distillation. The methodology uses HRWS with expert-tuned parameters to generate high-quality training targets, enabling neural networks to learn expert processing decisions. The model features a 51-channel input structure (50 amplitude offsets plus offset-density mask) and incorporates advanced attention mechanisms (CBAM, ASPP) for residual prediction. For industrial deployment, models are exported to ONNX format and deployed in distributed C++ environments with GPU acceleration. Experimental results demonstrate 15–20× speed improvement over classical methods while maintaining expert-level quality on diverse datasets without manual parameter tuning.

View free PDFSource page

Related papers

openalexZenodo (CERN European Organization for Nuclear Research)2026-07-24

Automatic Coronary Artery Segmentation in X-ray Angiograms

Filip Filipović, Tijana Geroski, N Filipovic

Coronary artery disease is one of the leading causes of morbidity and mortality worldwide, with X-ray coronary angiography serving as the clinical gold standard for diagnosis and intervention planning. Accurate segmentation of coronary arteries is essential for quantitative analy…

View free PDFSource page
openalexZenodo (CERN European Organization for Nuclear Research)2026-07-23

Dataset for "Deep learning models for estimating volume and Lorey's height across Nordic countries using optical and SAR satellite images" article

Zsófia Koma, Oleg Antropov, Jukka Miettinen, Johannes Breidenbach

This repository contains the data products and code required to reproduce the results presented in the article "Deep Learning Models for Estimating Volume and Lorey's Height Across Nordic Countries Using Optical and SAR Satellite Images". The study investigates the use of U-Net d…

View free PDFSource page
openalexZenodo (CERN European Organization for Nuclear Research)2026-07-24

Implementation source code of Neural Computing and Applications (2023) 35:19279–19295 https://doi.org/10.1007/s00521-023-08717-4 ,- volV) (0123456789().,-volV)(0123456789().

Javeria Amin

Source Code for: Detection of Brain Space-Occupying Lesions Using Quantum Machine Learning", "description": "Source code for the four-phase deep-learning pipeline presented in Amin, J., Anjum, M. A., Gul, N., & Sharif, M. (2023), 'Detection of brain space-occupying lesions using…

View free PDFSource page
openalexZenodo (CERN European Organization for Nuclear Research)2026-07-26

Electrical resistivity tomography surveys, trained physics-informed neural network models and code for amortized ERT inversion along Route Regionale 707, Moroccan Middle Atlas

Rajae Ajana

This deposit contains the field data, synthetic training datasets, trained network weights and analysis code supporting the article "Physics-informed neural network inversion of electrical resistivity tomography data: amortized optimization with field validation in the Moroccan M…

View free PDFSource page
openalexZenodo (CERN European Organization for Nuclear Research)2026-07-24

AI-Powered Fault Detection and Interpretation: From Neural Networks to Ready-to-Use Fault Surfaces

Alexander Shcherbina, Petr Popov, Ruslan Peisakhov, Yulia Sherman, Alex Berkovich

We present a comprehensive automated solution for 3D seismic fault detection and interpretation that combines deep learning with advanced geometric post-processing. The method integrates a 3D U-Net neural network trained on synthetic data with normalized distance function targets…

View free PDFSource page
openalexZenodo (CERN European Organization for Nuclear Research)2026-07-25

QSM-CI method: BFRnet (v1)

Xuanyu Zhu, Yang Gao, Feng Liu, Stuart Crozier, Hongfu Sun

Deep-learning background field removal (BFRnet): a 3D dual-frequency octave-convolution U-net trained to predict the background field of the brain — including brains with significant pathological susceptibility sources (haemorrhage, calcification). Consumes the total field (ppm)…

View free PDFSource page