Francesco Picetti

Bio: Francesco Picetti is an academic researcher from Instituto Politécnico Nacional. The author has contributed to research in topics: Ground-penetrating radar & Autoencoder. The author has an hindex of 4, co-authored 15 publications receiving 55 citations. Previous affiliations of Francesco Picetti include Polytechnic University of Milan.

Deep Prior-Based Unsupervised Reconstruction of Irregularly Sampled Seismic Data

Abstract: 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.

Landmine Detection Using Autoencoders on Multipolarization GPR Volumetric Data

Abstract: Buried landmines and unexploded remnants of war are a constant threat for the population of many countries that have been hit by wars in the past years. The huge amount of casualties has been a strong motivation for the research community toward the development of safe and robust techniques designed for landmine clearance. Nonetheless, being able to detect and localize buried landmines with high precision in an automatic fashion is still considered a challenging task due to the many different boundary conditions that characterize this problem (e.g., several kinds of objects to detect, different soils and meteorological conditions, etc.). In this article, we propose a novel technique for buried object detection tailored to unexploded landmine discovery. The proposed solution exploits a specific kind of convolutional neural network (CNN) known as autoencoder to analyze volumetric data acquired with ground penetrating radar (GPR) using different polarizations. This method works in an anomaly detection framework, indeed we only train the autoencoder on GPR data acquired on landmine-free areas. The system then recognizes landmines as objects that are dissimilar to the soil used during the training step. Experiments conducted on real data show that the proposed technique requires little training and no ad hoc data preprocessing to achieve accuracy higher than 93% on challenging data sets.

Convolutional Autoencoder for Landmine Detection on GPR Scans

Abstract: Buried unexploded landmines are a serious threat in many countries all over the World. As many landmines are nowadays mostly plastic made, the use of ground penetrating radar (GPR) systems for their detection is gaining the trend. However, despite several techniques have been proposed, a safe automatic solution is far from being at hand. In this paper, we propose a landmine detection method based on convolutional autoencoder applied to B-scans acquired with a GPR. The proposed system leverages an anomaly detection pipeline: the autoencoder learns a description of B-scans clear of landmines, and detects landmine traces as anomalies. In doing so, the autoencoder never uses data containing landmine traces at training time. This allows to avoid making strong assumptions on the kind of landmines to detect, thus paving the way to detection of novel landmine models.

Seismic image processing through the generative adversarial network

Abstract: 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 reas...

A generative adversarial network for seismic imaging applications

Abstract: 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.

DNNs as Applied to Electromagnetics, Antennas, and Propagation—A Review

Abstract: A review of the most recent advances in deep learning (DL) as applied to electromagnetics (EM), antennas, and propagation is provided. It is aimed at giving the interested readers and practitioners in EM and related applicative fields some useful insights on the effectiveness and potentialities of deep neural networks (DNNs) as computational tools with unprecedented computational efficiency. The range of considered applications includes forward/inverse scattering, direction-of-arrival estimation, radar and remote sensing, and multi-input/multi-output systems. Appealing DNN-based solutions concerned with localization, human behavior monitoring, and EM compatibility are reported as well. Some final remarks are drawn along with the indications on future trends according to the authors’ viewpoint.

A fully unsupervised and highly generalized deep learning approach for random noise suppression

Abstract: In this study, we proposed a deep learning algorithm (PATCHUNET) to suppress random noise and preserve the coherent seismic signal. The input data are divided into several patches, and each patch is encoded to extract the meaningful features. Following this, the extracted features are decompressed to retrieve the seismic signal. Skip connections are used between the encoder and decoder parts, allowing the proposed algorithm to extract high‐order features without losing important information. Besides, dropout layers are used as regularization layers. The dropout layers preserve the most meaningful features belonging to the seismic signal and discard the remaining features. The proposed algorithm is an unsupervised approach that does not require prior information about the clean signal. The input patches are divided into 80% for training and 20% for testing. However, it is interesting to find that the proposed algorithm can be trained with only 30% of the input patches with an effective denoising performance. Four synthetic and four field examples are used to evaluate the proposed algorithm performance, and compared to the f−x deconvolution and the f−x singular spectrum analysis. The results indicate the ability of the proposed algorithm in attenuating the random noise and preserving the seismic signal effectively despite the existence of a large amount of random noise, for example, when the input signal‐to‐noise ratio is as low as −14.2 dB.

Seismic data interpolation using deep learning with generative adversarial networks

Abstract: We propose an algorithm for seismic trace interpolation using generative adversarial networks, a type of deep neural network. The method extracts feature vectors from the training data using self‐learning and does not require any pre‐processing to create the training labels. The algorithm also does not make any prior explicit assumptions about linearity of seismic events or sparsity of the data, which are often required in the traditional interpolation methods. We create the training labels by removing traces from different receiver indices of the original datasets to simulate the effect of missing traces. We adopt the framework of the generative adversarial networks to train the network and add additional loss functions to regularize the model. Numerical examples using land and marine field datasets demonstrate the validity and effectiveness of the proposed approach. With minimal computational burden and proper training, the proposed method can be applied to three‐dimensional seismic datasets to achieve accurate interpolation results.