Download to read offline
Uploaded byDing Li
Seismic data analysis with u net
This document discusses using a U-Net neural network model to analyze seismic data and identify salt deposits beneath the Earth's surface. It provides details on the U-Net architecture, including convolutional and max pooling layers to capture context, upsampling and concatenation to combine context with local details. The model was trained on 3,200 seismic images to segment salt deposits, achieving an IOU of 72%. Several modifications were made to improve performance, such as adding batch normalization and increasing depth, resulting in an IOU of 77%.
![RTP_AR_Basic_Learners' Workbook_KS2 [FOR REPRODUCTION] (1).pdf](https://cdn.slidesharecdn.com/ss_thumbnails/rtparbasiclearnersworkbookks2forreproduction1-251016024943-e51a16ac-thumbnail.jpg?width=600ounds&width=560&fit=bounds)
Seismic data analysis with u net
- 1. Salt Identification Seismic DataAnalysis with Neural Network Ding Li 2018.08
- 2. 2 When a seismicwave travels through an interface of two materials, a reflection wave will be generated. https://en.wikipedia.org/wiki/Reflection_seismology http://geologylearn.blogspot.com/2015/06/marine-and-land-seismic-aquisition.html
- 4. 4 128 x 128 16 1 16 32 32 16 3264 64 64 128 128 128 256 256 128 128 128 128 64 64 64 64 32 32 32 32 16 16 16 1 128 x 128 Input Image File Output Segmentation Map conv 3x3, ReLU copy & concatenate max pool 2x2 up-conv 2x2 conv 1x1 Softmax Learn Context Structure Combine Context & Local Details High Resolution Shallow Low Resolution Deep High Resolution Shallow Low Resolution Deep U-Net: Convolutional Networks for Biomedical Image Segmentation Total Parameters: 2,158,417
- 5. 5 TGS Salt IdentificationChallenge Segment salt deposits beneath the Earth's surface Training images: 3,200 Evaluation images: 800 IOU = 72% After 40 epochs of training (2 hours on Kaggle). Grand Truth Prediction False Negative (FN) False Positive (FP) True Positive (TP) 𝐼𝑂𝑈 = 𝐼𝑛𝑡𝑒𝑟𝑠𝑒𝑐𝑡𝑖𝑜𝑛 𝑈𝑛𝑖𝑜𝑛 = 𝑇𝑃 𝑇𝑃 + 𝐹𝑁 + 𝐹𝑃 Image Mask Predict
- 7. 7 Why Does BatchNorm Work? (C2W3L06) Batch Normalization: Accelerating Deep Network Training b y Reducing Internal Covariate Shift Input Covariant Shift Normalize the Input to Improve Prediction Normalize each hidden layer, so there is less covariant shift for next layer
- 8. 8 • Without domainknowledge, U-NET can do a reasonable job to analyze seismic data. • Analyzing seismic data ourselves is using our own brain. • Appling neural networks is like building a new brain, then train the brain with data to do the job. • IOU was increased to 77% with the following steps • Added a batch normalization layer after each convolutional layer • Replaced Conv2DTranspose with UpSampling2D + Conv2D, making up-sampling more smooth. • Increased the depth to 5 with a new layer at the bottom of U-Net: 4x4 (512 filters) • How to improve further? • Apply conditional random fields (CRF) to enhance pattern recognition, check this paper. Python code of the project at kaggle: https://www.kaggle.com/dingli/seismic-data-analysis-with-u-net