# Deep Compressed Learning for 3D Seismic Inversion

arXiv: https://arxiv.org/abs/2311.00107 · HTML: https://mlsanity.com/paper/2311.00107/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 7 of its 26 references that are in the map.
- First posted on arXiv 31 Oct 2023.

## Builds on (most cited first, 7 of 7)

- [U-Net: Convolutional Networks for Biomedical Image Segmentation](https://mlsanity.com/paper/1505.04597/) (2015)
- [Deep-learning inversion: a next generation seismic velocity-model building method](https://mlsanity.com/paper/1902.06267/) (2019)
- [InversionNet3D: Efficient and Scalable Learning for 3D Full Waveform Inversion](https://mlsanity.com/paper/2103.14158/) (2021)
- [Compressed Learning: A Deep Neural Network Approach](https://mlsanity.com/paper/1610.09615/) (2016)
- [TransCL: Transformer Makes Strong and Flexible Compressive Learning](https://mlsanity.com/paper/2207.11972/) (2022)
- [Encoder-Decoder Architecture for 3D Seismic Inversion](https://mlsanity.com/paper/2207.14789/) (2022)
- [Minimod: A Finite Difference solver for Seismic Modeling](https://mlsanity.com/paper/2007.06048/) (2020)

## Built on by (most cited first, 2 of 2)

- [Physics-guided Full Waveform Inversion using Encoder-Solver Convolutional Neural Networks](https://mlsanity.com/paper/2405.17696/) (2024)
- [Accelerated Full Waveform Inversion by Deep Compressed Learning](https://mlsanity.com/paper/2601.01268/) (2026)

## Abstract (arXiv)

We consider the problem of 3D seismic inversion from pre-stack data using a very small number of seismic sources. The proposed solution is based on a combination of compressed-sensing and machine learning frameworks, known as compressed-learning. The solution jointly optimizes a dimensionality reduction operator and a 3D inversion encoder-decoder implemented by a deep convolutional neural network (DCNN). Dimensionality reduction is achieved by learning a sparse binary sensing layer that selects a small subset of the available sources, then the selected data is fed to a DCNN to complete the regression task. The end-to-end learning process provides a reduction by an order-of-magnitude in the number of seismic records used during training, while preserving the 3D reconstruction quality comparable to that obtained by using the entire dataset.