This work introduces composite functions to compute distortion in volumetric seismic data. Several loss functions, such as those based on L p -functions, ignore the structure of 3-d seismic data, treating it as a unid...
详细信息
This work introduces composite functions to compute distortion in volumetric seismic data. Several loss functions, such as those based on L p -functions, ignore the structure of 3-d seismic data, treating it as a unidimensional vector. Alternatively, applying distinct functions in each axis, properly designed for dimension reduction, can evaluate seismic data error according to its unity and magnitude. We thus propose a novel multidimensional composite loss function to evaluate through dimension reduction, suitable for seismic data compression within a method named3dSC-GAN. It replaces the usual peak signal-to-noise ratio (PSNR) metric as the distortion function. An extensive study is conducted to analyze potential combinations of functions for the 3-dpoststack seismic data compression problem. Results indicate that the new function contributes to improve the neural network learning step. Our method provides superior reconstructions, both quantitatively and qualitatively, when compared to the PSNR metric.
We approach the problem of 3-dpoststack seismic data compression by training a model based on a deep autoencoder. Our network architecture is trained to consider the similarity between 3-d seismic sections drawn from...
详细信息
We approach the problem of 3-dpoststack seismic data compression by training a model based on a deep autoencoder. Our network architecture is trained to consider the similarity between 3-d seismic sections drawn from one or multiple seismic volumes. A whole seismic volume is compressed with the latent representations of each of its composing volumetric sections. The goal is to compress the seismic data at very low bit rates with high-quality reconstruction. Our model is suitable for training general compressors from multiple seismic surveys or for specialized compression of a single seismic volume. Results show that our method can compress seismic data with extremely low bit rates, below 0.3 bits-per-voxel (bpv) while yielding peak signal-to-noise ratio (PSNR) values over 40 dB.
This work presents a method for volumetric seismic data compression by coupling a 3-d convolution-based autoencoder to a generative adversarial network (GAN). The main challenge of 3-d convolutional autoencoders for d...
详细信息
This work presents a method for volumetric seismic data compression by coupling a 3-d convolution-based autoencoder to a generative adversarial network (GAN). The main challenge of 3-d convolutional autoencoders for data compression is how to fully exploit volumetric redundancy while keeping reasonable latent representation dimensions. Our method is based on a convolutional neural network for seismic data compression called3dSC. Its encoder anddecoder use 3-d convolutions and are connected by a latent representation with the same dimensions as its 2-d network counterparts. Our main hypothesis is that the 3dSC architecture can be improved by adversarial training. We, thus, propose a new 3-d-based seismic data compression method (3dSC-GAN) by coupling the 3dSC network to a GAN. The seismic datadecoder is used as a generator of poststackdata that are integrated with a discriminator module to better exploit 3-d redundancy. Results show that our method outperforms previous seismic data compression methods for very low target bit rates, increasing the peak signal-to-noise ratio (PSNR) with fairly high visual quality.
暂无评论