Semi-Supervised Segmentation of Salt Bodies in Seismic Images using an Ensemble of Convolutional Neural Networks (2019-04-09T00:00:00.000000Z)

TL;DR

A semi-supervised method for segmentation (delineation) of salt bodies in seismic images which utilizes unlabeled data for multi-round self-training and outperforms state-of-the-art on the TGS Salt Identification Challenge dataset and is ranked the first among the 3234 competing methods.

Abstract

Seismic image analysis plays a crucial role in a wide range of industrial applications and has been receiving significant attention. One of the essential challenges of seismic imaging is detecting subsurface salt structure which is indispensable for identification of hydrocarbon reservoirs and drill path planning. Unfortunately, exact identification of large salt deposits is notoriously difficult and professional seismic imaging often requires expert human interpretation of salt bodies. Convolutional neural networks (CNNs) have been successfully applied in many fields, and several attempts have been made in the field of seismic imaging. But the high cost of manual annotations by geophysics experts and scarce publicly available labeled datasets hinder the performance of the existing CNN-based methods. In this work, we propose a semi-supervised method for segmentation (delineation) of salt bodies in seismic images which utilizes unlabeled data for multi-round self-training. To reduce error amplification during self-training we propose a scheme which uses an ensemble of CNNs. We show that our approach outperforms state-of-the-art on the TGS Salt Identification Challenge dataset and is ranked the first among the 3234 competing methods.

Authors

Artsiom Sanakoyeu

2 papers

Yauhen Babakhin

1 papers

Hirotoshi Kitamura

1 papers

TL;DR

Abstract

Authors

References60 items

Volume 1

A New Ensemble Self-labeled Semi-supervised Algorithm

Multi-resolution neural networks for tracking seismic horizons from few training images

Automatic Seismic Salt Interpretation with Deep Convolutional Neural Networks

Automatic salt deposits segmentation: A deep learning approach

Albumentations: fast and flexible image augmentations

Multi-attribute k-means clustering for salt-boundary delineation from three-dimensional seismic data

Seismic facies analysis using machine learning

Deep-learning seismic facies on state-of-the-art CNN architectures

Real-time seismic-image interpretation via deconvolutional neural network

Improving Spatiotemporal Self-Supervision by Deep Reinforcement Learning

An Ensemble SSL Algorithm for Efficient Chest X-Ray Image Classification

Convolutional neural networks for automated seismic interpretation

Automatic Salt Detection with Machine Learning

SFA-GTM: Seismic Facies Analysis Based on Generative Topographic Map and RBF

Pyramid Attention Network for Semantic Segmentation

Concurrent Spatial and Channel Squeeze & Excitation in Fully Convolutional Networks

Deep unsupervised learning of visual similarities

Progressive Semisupervised Learning of Multiple Classifiers

TernausNet: U-Net with VGG11 Encoder Pre-Trained on ImageNet for Image Segmentation

A semi-supervised generative framework with deep learning features for high-resolution remote sensing image scene classification

Deep Growing Learning

Unsupervised Representation Learning by Sorting Sequences

The Lovasz-Softmax Loss: A Tractable Surrogate for the Optimization of the Intersection-Over-Union Measure in Neural Networks

Deep Unsupervised Similarity Learning Using Partially Ordered Sets

Deep Neural Network Self-training Based on Unsupervised Learning and Dropout

Aggregated Residual Transformations for Deep Neural Networks

Methods to compute salt likelihoods and extract salt boundaries from 3D seismic images

CliqueCNN: Deep Unsupervised Exemplar Learning

SGDR: Stochastic Gradient Descent with Warm Restarts

Context Encoders: Feature Learning by Inpainting

Unsupervised Learning of Visual Representations by Solving Jigsaw Puzzles

Learning Representations for Automatic Colorization

Self-Trained LMT for Semisupervised Learning

Deep Residual Learning for Image Recognition

A comparison of classification techniques for seismic facies recognition

Unsupervised Visual Representation Learning by Context Prediction

U-Net: Convolutional Networks for Biomedical Image Segmentation

Self-labeled techniques for semi-supervised learning: taxonomy, software and empirical study

Semantic Image Segmentation with Deep Convolutional Nets and Fully Connected CRFs

Hypercolumns for object segmentation and fine-grained localization

Seismic imaging in and around salt bodies

Texture Attributes for Detecting Salt Bodies in Seismic Data

SETRED: Self-training with Editing

Unsupervised Word Sense Disambiguation Rivaling Supervised Methods

A texture-based approach to the segmentation of seismic images

Backpropagation Applied to Handwritten Zip Code Recognition

Segmentation models zoo in Keras

Pengfei Xiong

Pseudo-Label : The Simple and Efficient Semi-Supervised Learning Method for Deep Neural Networks

Learning Multiple Layers of Features from Tiny Images

Salt body segmentation with dip and frequency attributes

Semi-Supervised Learning Literature Survey

Accessed

Applied Geophysics: Contents

S UPPLEMENTARY M ATERIAL OF Self-Supervised Relative Depth Learning for Urban Scene Understanding

A New Ensemble Semi-supervised Self-labeled Algorithm

Applied Geophysics

The Lov´asz-Softmax loss: A tractable surrogate for the optimization of the intersection-over-union measure in neural networks

TGS salt identification challenge

Field of Study

Venue Information

Name

Type

URL

Alternate Names