Juan José Murillo-Fuentes

A. Delgado, L. Alba-Carcelén, J. J. Murillo-Fuentes

In the forensic studies of painting masterpieces, the analysis of the support is of major importance. For plain weave fabrics, the densities of vertical and horizontal threads are used as main features, while angle deviations from the vertical and horizontal axis are also of help. These features can be studied locally through the canvas. In this work, deep learning is proposed as a tool to perform these local densities and angle studies. We trained the model with samples from 36 paintings by Velázquez, Rubens or Ribera, among others. The data preparation and augmentation are dealt with at a first stage of the pipeline. We then focus on the supervised segmentation of crossing points between threads. The U-Net with inception and Dice loss are presented as good choices for this task. Densities and angles are then estimated based on the segmented crossing points. We report test results of the analysis of a few canvases and a comparison with methods in the frequency domain, widely used in this problem. We concluded that this new approach succeeds in some cases where the frequency analysis tools fail, while improving the results in others. Besides, our proposal does not need the labeling of part of the to-be-processed image. As case studies, we apply this novel algorithm to the analysis of two pairs of canvases by Velázquez and Murillo, to conclude that the fabrics used came from the same roll.

A. D. Bejarano, Juan J. Murillo-Fuentes, Laura Alba-Carcelen

In this work, the authors develop regression approaches based on deep learning to perform thread density estimation for plain weave canvas analysis. Previous approaches were based on Fourier analysis, which is quite robust for some scenarios but fails in some others, in machine learning tools, that involve pre-labeling of the painting at hand, or the segmentation of thread crossing points, that provides good estimations in all scenarios with no need of pre-labeling. The segmentation approach is time-consuming as the estimation of the densities is performed after locating the crossing points. In this novel proposal, we avoid this step by computing the density of threads directly from the image with a regression deep learning model. We also incorporate some improvements in the initial preprocessing of the input image with an impact on the final error. Several models are proposed and analyzed to retain the best one. Furthermore, we further reduce the density estimation error by introducing a semi-supervised approach. The performance of our novel algorithm is analyzed with works by Ribera, Velázquez, and Poussin where we compare our results to the ones of previous approaches. Finally, the method is put into practice to support the change of authorship or a masterpiece at the Museo del Prado.

Juan José Murillo-Fuentes, Pablo M. Olmos, Laura Alba-Carcelén

The study of canvas fabrics in works of art is a crucial tool for authentication, attribution and conservation. Traditional methods are based on thread density map matching, which cannot be applied when canvases do not come from contiguous positions on a roll. This paper presents a novel approach based on deep learning to assess the similarity of textiles. We introduce an automatic tool that evaluates the similarity between canvases without relying on thread density maps. A Siamese deep learning model is designed and trained to compare pairs of images by exploiting the feature representations learned from the scans. In addition, a similarity estimation method is proposed, aggregating predictions from multiple pairs of cloth samples to provide a robust similarity score. Our approach is applied to canvases from the Museo Nacional del Prado, corroborating the hypothesis that plain weave canvases, widely used in painting, can be effectively compared even when their thread densities are similar. The results demonstrate the feasibility and accuracy of the proposed method, opening new avenues for the analysis of masterpieces.

José Carlos Aradillas, Juan José Murillo-Fuentes, Pablo M. Olmos

In this paper, we face the problem of offline handwritten text recognition (HTR) in historical documents when few labeled samples are available and some of them contain errors in the train set. Three main contributions are developed. First we analyze how to perform transfer learning (TL) from a massive database to a smaller historical database, analyzing which layers of the model need a fine-tuning process. Second, we analyze methods to efficiently combine TL and data augmentation (DA). Finally, an algorithm to mitigate the effects of incorrect labelings in the training set is proposed. The methods are analyzed over the ICFHR 2018 competition database, Washington and Parzival. Combining all these techniques, we demonstrate a remarkable reduction of CER (up to 6% in some cases) in the test set with little complexity overhead.

Rafael Boloix-Tortosa, Juan José Murillo-Fuentes, Sotirios A. Tsaftaris

We propose a novel adaptive kernel based regression method for complex-valued signals: the generalized complex-valued kernel least-mean-square (gCKLMS). We borrow from the new results on widely linear reproducing kernel Hilbert space (WL-RKHS) for nonlinear regression and complex-valued signals, recently proposed by the authors. This paper shows that in the adaptive version of the kernel regression for complex-valued signals we need to include another kernel term, the so-called pseudo-kernel. This new solution is endowed with better representation capabilities in complex-valued fields, since it can efficiently decouple the learning of the real and the imaginary part. Also, we review previous realizations of the complex KLMS algorithm and its augmented version to prove that they can be rewritten as particular cases of the gCKLMS. Furthermore, important conclusions on the kernels design are drawn that help to greatly improve the convergence of the algorithms. In the experiments, we revisit the nonlinear channel equalization problem to highlight the better convergence of the gCKLMS compared to previous solutions. Also, the flexibility of the proposed generalized approach is tested in a second experiment with non-independent real and imaginary parts. The results illustrate the significant performance improvements of the gCKLMS approach when the complex-valued signals have different properties for the real and imaginary parts.

Marton Havasi, José Miguel Hernández-Lobato, Juan José Murillo-Fuentes

Deep Gaussian Processes (DGPs) are hierarchical generalizations of Gaussian Processes that combine well calibrated uncertainty estimates with the high flexibility of multilayer models. One of the biggest challenges with these models is that exact inference is intractable. The current state-of-the-art inference method, Variational Inference (VI), employs a Gaussian approximation to the posterior distribution. This can be a potentially poor unimodal approximation of the generally multimodal posterior. In this work, we provide evidence for the non-Gaussian nature of the posterior and we apply the Stochastic Gradient Hamiltonian Monte Carlo method to generate samples. To efficiently optimize the hyperparameters, we introduce the Moving Window MCEM algorithm. This results in significantly better predictions at a lower computational cost than its VI counterpart. Thus our method establishes a new state-of-the-art for inference in DGPs.

Marton Havasi, José Miguel Hernández-Lobato, Juan José Murillo-Fuentes

Deep Gaussian Processes (DGP) are hierarchical generalizations of Gaussian Processes (GP) that have proven to work effectively on a multiple supervised regression tasks. They combine the well calibrated uncertainty estimates of GPs with the great flexibility of multilayer models. In DGPs, given the inputs, the outputs of the layers are Gaussian distributions parameterized by their means and covariances. These layers are realized as Sparse GPs where the training data is approximated using a small set of pseudo points. In this work, we show that the computational cost of DGPs can be reduced with no loss in performance by using a separate, smaller set of pseudo points when calculating the layerwise variance while using a larger set of pseudo points when calculating the layerwise mean. This enabled us to train larger models that have lower cost and better predictive performance.

Francisco J. Simois, Juan J. Murillo-Fuentes

A routine task for art historians is painting diagnostics, such as dating or attribution. Signal processing of the X-ray image of a canvas provides useful information about its fabric. However, previous methods may fail when very old and deteriorated artworks or simply canvases of small size are studied. We present a new framework to analyze and further characterize the paintings from their radiographs. First, we start from a general analysis of lattices and provide new unifying results about the theoretical spectra of weaves. Then, we use these results to infer the main structure of the fabric, like the type of weave and the thread densities. We propose a practical estimation of these theoretical results from paintings with the averaged power spectral density (PSD), which provides a more robust tool. Furthermore, we found that the PSD provides a fingerprint that characterizes the whole canvas. We search and discuss some distinctive features we may find in that fingerprint. We apply these results to several masterpieces of the 17th and 18th centuries from the Museo Nacional del Prado to show that this approach yields accurate results in thread counting and is very useful for paintings comparison, even in situations where previous methods fail.

Rafael Boloix-Tortosa, Juan José Murillo-Fuentes, Irene Santos Velázquez et al.

Usually, complex-valued RKHS are presented as an straightforward application of the real-valued case. In this paper we prove that this procedure yields a limited solution for regression. We show that another kernel, here denoted as pseudo kernel, is needed to learn any function in complex-valued fields. Accordingly, we derive a novel RKHS to include it, the widely RKHS (WRKHS). When the pseudo-kernel cancels, WRKHS reduces to complex-valued RKHS of previous approaches. We address the kernel and pseudo-kernel design, paying attention to the kernel and the pseudo-kernel being complex-valued. In the experiments included we report remarkable improvements in simple scenarios where real a imaginary parts have different similitude relations for given inputs or cases where real and imaginary parts are correlated. In the context of these novel results we revisit the problem of non-linear channel equalization, to show that the WRKHS helps to design more efficient solutions.

Rafael Boloix-Tortosa, Eva Arias-de-Reyna, F. Javier Payan-Somet et al.

In this paper we propose a novel Bayesian solution for nonlinear regression in complex fields. Previous solutions for kernels methods usually assume a complexification approach, where the real-valued kernel is replaced by a complex-valued one. This approach is limited. Based on results in complex-valued linear theory and Gaussian random processes we show that a pseudo-kernel must be included. This is the starting point to develop the new complex-valued formulation for Gaussian process for regression (CGPR). We face the design of the covariance and pseudo-covariance based on a convolution approach and for several scenarios. Just in the particular case where the outputs are proper, the pseudo-kernel cancels. Also, the hyperparameters of the covariance {can be learnt} maximizing the marginal likelihood using Wirtinger's calculus and patterned complex-valued matrix derivatives. In the experiments included, we show how CGPR successfully solve systems where real and imaginary parts are correlated. Besides, we successfully solve the nonlinear channel equalization problem by developing a recursive solution with basis removal. We report remarkable improvements compared to previous solutions: a 2-4 dB reduction of the MSE with {just a quarter} of the training samples used by previous approaches.

Rafael Boloix-Tortosa, F. Javier Payán-Somet, Eva Arias-de-Reyna et al.

Complex-valued signals are used in the modeling of many systems in engineering and science, hence being of fundamental interest. Often, random complex-valued signals are considered to be proper. A proper complex random variable or process is uncorrelated with its complex conjugate. This assumption is a good model of the underlying physics in many problems, and simplifies the computations. While linear processing and neural networks have been widely studied for these signals, the development of complex-valued nonlinear kernel approaches remains an open problem. In this paper we propose Gaussian processes for regression as a framework to develop 1) a solution for proper complex-valued kernel regression and 2) the design of the reproducing kernel for complex-valued inputs, using the convolutional approach for cross-covariances. In this design we pay attention to preserve, in the complex domain, the measure of similarity between near inputs. The hyperparameters of the kernel are learned maximizing the marginal likelihood using Wirtinger derivatives. Besides, the approach is connected to the multiple output learning scenario. In the experiments included, we first solve a proper complex Gaussian process where the cross-covariance does not cancel, a challenging scenario when dealing with proper complex signals. Then we successfully use these novel results to solve some problems previously proposed in the literature as benchmarks, reporting a remarkable improvement in the estimation error.