Vincenzo Lipari

Seismic swell noise, often observed in marine seismic data, is characterised by high amplitude and low frequencies. This noise significantly hides useful signals, underscoring the importance of attenuating it in the processing pipeline for marine seismic data. To date, most deep learning methods proposed for swell noise suppression have focused on supervised paradigms, which require a large dataset of paired noisy and clean data to gain insight into signal features and swell noise statistics at training phase. In field data processing, however, obtaining generalizable training data often proves to be challenging. The inherent complexity and variability of the real world often make it difficult to acquire unbiased, noise-free datasets in the context of marine seismic data processing. To address this problem, we present a self-supervised deep learning method for suppressing swell noise even in the absence of access to clean seismic training data. First, we propose a strategy for the synthetic swell noise generation based on the amplitude, frequency and coherence features of noisy traces. We implement the Noisy-As-Clean (NAC) strategy, wherein either the original noisy seismic data or its reorganized variant acts as network’s target. Simultaneously, the network receives the observed noisy seismic data combined with simulated swell noise as its input. By leveraging these “noisy-noisy” pairs, we train a DnCNN network. Experimental evaluations conducted on both synthetic and field data demonstrate that, through the integration of swell noise simulation and the NAC strategy, the trained network consistently achieves superior denoising performance.

Seismic deblending is an ill-posed inverse problem that involves counteracting the effect of a blending matrix derived from the shots’ position and firing time. In this letter, we propose a seismic deblending method based on so-called deep preconditioners. A convolutional autoencoder (AE) is first trained in a patch-wise fashion to learn an effective sparse representation of the common receiver gathers (CRGs) we aim to reconstruct. Then, the decoder branch of the trained AE is used as a nonlinear preconditioner for the deblending problem. Particularly, to avoid the explicit creation of a training dataset, we suggest to use the common shot gathers (CSGs) of the blended dataset itself to train the AE network, as they are not affected by incoherent blending noise. Numerical examples on synthetic and field datasets demonstrate the effectiveness of the proposed method in comparison to significantly comparable techniques: a dictionary-learning-based deblending method; an end-to-end deblending convolution neutral network (CNN).

Irregularity and coarse spatial sampling of seismic data strongly affect the performances of processing and imaging algorithms. Therefore, interpolation is a usual preprocessing step in most of the processing workflows. In this work, we propose a seismic data interpolation method based on the deep prior paradigm: an ad hoc convolutional neural network is used as a prior to solve the interpolation inverse problem, avoiding any costly and prone-to-overfitting training stage. In particular, the proposed method leverages a multiresolution U-Net with 3-D convolution kernels exploiting correlations in cubes of seismic data, at different scales in all directions. Numerical examples on different corrupted synthetic and field data sets show the effectiveness and promising features of the proposed approach.

The advent of new deep-learning and machine-learning paradigms enables the development of new solutions to tackle the challenges posed by new geophysical imaging applications. For this reason, convolutional neural networks (CNNs) have been deeply investigated as novel tools for seismic image processing. In particular, we have studied a specific CNN architecture, the generative adversarial network (GAN), through which we process seismic migrated images to obtain different kinds of output depending on the application target defined during training. We have developed two proof-of-concept applications. In the first application, a GAN is trained to turn a low-quality migrated image into a high-quality one, as if the acquisition geometry was much more dense than in the input. In the second example, the GAN is trained to turn a migrated image into the respective deconvolved reflectivity image. The effectiveness of the investigated approach is validated by means of tests performed on synthetic examples.

A common issue of seismic data analysis consists in the lack of regular and densely sampled seismic traces. This problem is commonly tackled by rank optimization or statistical features learning algorithms, which allow interpolation and denoising of corrupted data. In this paper, we propose a completely novel approach for reconstructing missing traces of pre-stack seismic data, taking inspiration from computer vision and image processing latest developments. More specifically, we exploit a specific kind of convolutional neural networks known as convolutional autoencoder. We illustrate the advantages of using deep learning strategies with respect to state-of-the-art by comparing the achieved results over a well-known seismic dataset. Presentation Date: Wednesday, October 17, 2018 Start Time: 1:50:00 PM Location: 204C (Anaheim Convention Center) Presentation Type: Oral

The new challenges of geophysical imaging applications ask for new methodologies going beyond the standard and well established techniques. In this work we propose a novel tool for seismic imaging applications based on recent advances in deep neural networks. Specifically, we use a generative adversarial network (GAN) to process seismic migrated images in order to potentially obtain different kinds of outputs depending on the application target at training stage. We demonstrate the promising features of this tool through a couple of synthetic examples. In the first example, the GAN is trained to turn a low-quality migrated image into a high-quality one, as if the acquisition geometry were much more dense than in the input. In the second example, the GAN is trained to turn a migrated image into the respective deconvolved reflectivity image. Presentation Date: Wednesday, October 17, 2018 Start Time: 9:20:00 AM Location: Poster Station 1 Presentation Type: Poster

Although, with the usual seismic acquisition, much of the information about subsurface parameters is contained in the reflections data, many of the current Full Waveform Inversion applications still focus on the use of the transmitted energy (i.e. refractions). In order to reduce the dependency on refractions, in this work we propose to pre-process the recorded and the simulated signals with an exponential function in order to be able to exploit the Normalized Integration Method also for reflection data. Tests on synthetic data shows that the resulting new cost function is more robust with respect to the cycle skip compared to the standard least squares cost function. We show the results of the proposed method on two simple synthetic datasets and on a more realistic one characterized by complex geology and strong velocity contrasts Presentation Date: Thursday, October 18, 2018 Start Time: 8:30:00 AM Location: 207C (Anaheim Convention Center) Presentation Type: Oral

Due to the increasingly unbridled practice of sharing visual content on the web, tracing back past history of uploaded images is getting far from being an easy task. Nonetheless, forensic analysts might be interested in probing digital history of content published on the web to assess its authenticity. In this vein, a possible indicator of image integrity is the number of JPEG compressions a picture underwent. As a matter of fact, JPEG compression is typically operated first at image inception time directly on the acquisition device. Then, it is customary re-applied every time an image is manipulated or shared through social media. For this reason, the more the applied JPEG compressions, the more the likelihood that an image underwent some editing. In this work, we propose an algorithm to detect multiple JPEG compressions, specifically up to four coding cycles. This approach leverages the Task-driven Non-negative Matrix Factorization (TNMF) model, fed with histograms of the Discrete Cosine Transform (DCT) of the image under analysis. Experimental results show the effectiveness of the method if compared with the state-of-the-art, confirming this strategy as a viable solution for detecting multiple JPEG compressions.

Over the years, the forensic community has developed a series of very accurate camera attribution algorithms enabling to detect which device has been used to acquire an image with outstanding results. Many of these methods are based on photo response non uniformity (PRNU) that allows tracing back a picture to the camera used to shoot it. However, when privacy is required, it would be desirable to anonymize photos, unlinking them from their specific device. This paper investigates a new and alternative approach to image anonymization task. The proposed method leverages image inpainting described as an inverse regularized problem, and does not need any priors about the PRNU to remove. Specifically, we show how PRNU pattern can be strongly attenuated by reconstructing each pixel of an image from its neighbors, only slightly affecting visual quality. Results confirm this approach as a viable alternative solution for image anonymization.