arXiv subjects

Antonio Ortega

Publications and source records attributed to Antonio Ortega.

At least 91 records · Page 5Linked to original sources

arXiv2021

Multi-resolution intra-predictive coding of 3D point cloud attributes

We propose an intra frame predictive strategy for compression of 3D point cloud attributes. Our approach is integrated with the region adaptive graph Fourier transform (RAGFT), a multi-resolution transform formed by a composition of localized block transforms, which produces a set of low pass (approximation) and high pass (detail) coefficients at multiple resolutions. Since the transform operations are spatially localized, RAGFT coefficients at a given resolution may still be correlated. To exploit this phenomenon, we propose an intra-prediction strategy, in which decoded approximation coefficients are used to predict uncoded detail coefficients. The prediction residuals are then quantized and entropy coded. For the 8i dataset, we obtain gains up to 0.5db as compared to intra predicted point cloud compresion based on the region adaptive Haar transform (RAHT).

eess.IV↗

arXiv2021

Covariance Matrix Estimation with Non Uniform and Data Dependent Missing Observations

In this paper we study covariance estimation with missing data. We consider missing data mechanisms that can be independent of the data, or have a time varying dependency. Additionally, observed variables may have arbitrary (non uniform) and dependent observation probabilities. For each mechanism, we construct an unbiased estimator and obtain bounds for the expected value of their estimation error in operator norm. Our bounds are equivalent, up to constant and logarithmic factors, to state of the art bounds for complete and uniform missing observations. Furthermore, for the more general non uniform and dependent cases, the proposed bounds are new or improve upon previous results. Our error estimates depend on quantities we call scaled effective rank, which generalize the effective rank to account for missing observations. All the estimators studied in this work have the same asymptotic convergence rate (up to logarithmic factors).

math.ST↗

arXiv2021

DCT and DST Filtering with Sparse Graph Operators

Graph filtering is a fundamental tool in graph signal processing. Polynomial graph filters (PGFs), defined as polynomials of a fundamental graph operator, can be implemented in the vertex domain, and usually have a lower complexity than frequency domain filter implementations. In this paper, we focus on the design of filters for graphs with graph Fourier transform (GFT) corresponding to a discrete trigonometric transform (DTT), i.e., one of 8 types of discrete cosine transforms (DCT) and 8 discrete sine transforms (DST). In this case, we show that multiple sparse graph operators can be identified, which allows us to propose a generalization of PGF design: multivariate polynomial graph filter (MPGF). First, for the widely used DCT-II (type-2 DCT), we characterize a set of sparse graph operators that share the DCT-II matrix as their common eigenvector matrix. This set contains the well-known connected line graph. These sparse operators can be viewed as graph filters operating in the DCT domain, which allows us to approximate any DCT graph filter by a MPGF, leading to a design with more degrees of freedom than the conventional PGF approach. Then, we extend those results to all of the 16 DTTs as well as their 2D versions, and show how their associated sets of multiple graph operators can be determined. We demonstrate experimentally that ideal low-pass and exponential DCT/DST filters can be approximated with higher accuracy with similar runtime complexity. Finally, we apply our method to transform-type selection in a video codec, AV1, where we demonstrate significant encoding time savings, with a negligible compression loss.

eess.SP↗

arXiv2021

Learning Sparse Graph Laplacian with K Eigenvector Prior via Iterative GLASSO and Projection

Learning a suitable graph is an important precursor to many graph signal processing (GSP) pipelines, such as graph spectral signal compression and denoising. Previous graph learning algorithms either i) make some assumptions on connectivity (e.g., graph sparsity), or ii) make simple graph edge assumptions such as positive edges only. In this paper, given an empirical covariance matrix $\bar{C}$ computed from data as input, we consider a structural assumption on the graph Laplacian matrix $L$: the first $K$ eigenvectors of $L$ are pre-selected, e.g., based on domain-specific criteria, such as computation requirement, and the remaining eigenvectors are then learned from data. One example use case is image coding, where the first eigenvector is pre-chosen to be constant, regardless of available observed data. We first prove that the subspace of symmetric positive semi-definite (PSD) matrices $H_{u}^+$ with the first $K$ eigenvectors being $\{u_k\}$ in a defined Hilbert space is a convex cone. We then construct an operator to project a given positive definite (PD) matrix $L$ to $H_{u}^+$, inspired by the Gram-Schmidt procedure. Finally, we design an efficient hybrid graphical lasso/projection algorithm to compute the most suitable graph Laplacian matrix $L^* \in H_{u}^+$ given $\bar{C}$. Experimental results show that given the first $K$ eigenvectors as a prior, our algorithm outperforms competing graph learning schemes using a variety of graph comparison metrics.

eess.SP↗

arXiv2021

Spatio-Temporal Graph Scattering Transform

Although spatio-temporal graph neural networks have achieved great empirical success in handling multiple correlated time series, they may be impractical in some real-world scenarios due to a lack of sufficient high-quality training data. Furthermore, spatio-temporal graph neural networks lack theoretical interpretation. To address these issues, we put forth a novel mathematically designed framework to analyze spatio-temporal data. Our proposed spatio-temporal graph scattering transform (ST-GST) extends traditional scattering transforms to the spatio-temporal domain. It performs iterative applications of spatio-temporal graph wavelets and nonlinear activation functions, which can be viewed as a forward pass of spatio-temporal graph convolutional networks without training. Since all the filter coefficients in ST-GST are mathematically designed, it is promising for the real-world scenarios with limited training data, and also allows for a theoretical analysis, which shows that the proposed ST-GST is stable to small perturbations of input signals and structures. Finally, our experiments show that i) ST-GST outperforms spatio-temporal graph convolutional networks by an increase of 35% in accuracy for MSR Action3D dataset; ii) it is better and computationally more efficient to design the transform based on separable spatio-temporal graphs than the joint ones; and iii) the nonlinearity in ST-GST is critical to empirical performance.

eess.SP↗

arXiv2020

Representing Deep Neural Networks Latent Space Geometries with Graphs

Deep Learning (DL) has attracted a lot of attention for its ability to reach state-of-the-art performance in many machine learning tasks. The core principle of DL methods consists in training composite architectures in an end-to-end fashion, where inputs are associated with outputs trained to optimize an objective function. Because of their compositional nature, DL architectures naturally exhibit several intermediate representations of the inputs, which belong to so-called latent spaces. When treated individually, these intermediate representations are most of the time unconstrained during the learning process, as it is unclear which properties should be favored. However, when processing a batch of inputs concurrently, the corresponding set of intermediate representations exhibit relations (what we call a geometry) on which desired properties can be sought. In this work, we show that it is possible to introduce constraints on these latent geometries to address various problems. In more details, we propose to represent geometries by constructing similarity graphs from the intermediate representations obtained when processing a batch of inputs. By constraining these Latent Geometry Graphs (LGGs), we address the three following problems: i) Reproducing the behavior of a teacher architecture is achieved by mimicking its geometry, ii) Designing efficient embeddings for classification is achieved by targeting specific geometries, and iii) Robustness to deviations on inputs is achieved via enforcing smooth variation of geometry between consecutive latent spaces. Using standard vision benchmarks, we demonstrate the ability of the proposed geometry-based methods in solving the considered problems.

cs.LG↗

arXiv2020

Spectral folding and two-channel filter-banks on arbitrary graphs

In the past decade, several multi-resolution representation theories for graph signals have been proposed. Bipartite filter-banks stand out as the most natural extension of time domain filter-banks, in part because perfect reconstruction, orthogonality and bi-orthogonality conditions in the graph spectral domain resemble those for traditional filter-banks. Therefore, many of the well known orthogonal and bi-orthogonal designs can be easily adapted for graph signals. A major limitation is that this framework can only be applied to the normalized Laplacian of bipartite graphs. In this paper we extend this theory to arbitrary graphs and positive semi-definite variation operators. Our approach is based on a different definition of the graph Fourier transform (GFT), where orthogonality is defined with the respect to the Q inner product. We construct GFTs satisfying a spectral folding property, which allows us to easily construct orthogonal and bi-orthogonal perfect reconstruction filter-banks. We illustrate signal representation and computational efficiency of our filter-banks on 3D point clouds with hundreds of thousands of points.

eess.SP↗

arXiv2020

Graph-based Transforms for Video Coding

In many state-of-the-art compression systems, signal transformation is an integral part of the encoding and decoding process, where transforms provide compact representations for the signals of interest. This paper introduces a class of transforms called graph-based transforms (GBTs) for video compression, and proposes two different techniques to design GBTs. In the first technique, we formulate an optimization problem to learn graphs from data and provide solutions for optimal separable and nonseparable GBT designs, called GL-GBTs. The optimality of the proposed GL-GBTs is also theoretically analyzed based on Gaussian-Markov random field (GMRF) models for intra and inter predicted block signals. The second technique develops edge-adaptive GBTs (EA-GBTs) in order to flexibly adapt transforms to block signals with image edges (discontinuities). The advantages of EA-GBTs are both theoretically and empirically demonstrated. Our experimental results demonstrate that the proposed transforms can significantly outperform the traditional Karhunen-Loeve transform (KLT).

eess.IV↗

arXiv2020

Sampling Signals on Graphs: From Theory to Applications

The study of sampling signals on graphs, with the goal of building an analog of sampling for standard signals in the time and spatial domains, has attracted considerable attention recently. Beyond adding to the growing theory on graph signal processing (GSP), sampling on graphs has various promising applications. In this article, we review current progress on sampling over graphs focusing on theory and potential applications. Although most methodologies used in graph signal sampling are designed to parallel those used in sampling for standard signals, sampling theory for graph signals significantly differs from the theory of Shannon--Nyquist and shift-invariant sampling. This is due in part to the fact that the definitions of several important properties, such as shift invariance and bandlimitedness, are different in GSP systems. Throughout this review, we discuss similarities and differences between standard and graph signal sampling and highlight open problems and challenges.

eess.SP↗

arXiv2020

DeepNNK: Explaining deep models and their generalization using polytope interpolation

Modern machine learning systems based on neural networks have shown great success in learning complex data patterns while being able to make good predictions on unseen data points. However, the limited interpretability of these systems hinders further progress and application to several domains in the real world. This predicament is exemplified by time consuming model selection and the difficulties faced in predictive explainability, especially in the presence of adversarial examples. In this paper, we take a step towards better understanding of neural networks by introducing a local polytope interpolation method. The proposed Deep Non Negative Kernel regression (NNK) interpolation framework is non parametric, theoretically simple and geometrically intuitive. We demonstrate instance based explainability for deep learning models and develop a method to identify models with good generalization properties using leave one out estimation. Finally, we draw a rationalization to adversarial and generative examples which are inevitable from an interpolation view of machine learning.

cs.LG↗

arXiv2020

CoronaSurveys: Using Surveys with Indirect Reporting to Estimate the Incidence and Evolution of Epidemics

The world is suffering from a pandemic called COVID-19, caused by the SARS-CoV-2 virus. National governments have problems evaluating the reach of the epidemic, due to having limited resources and tests at their disposal. This problem is especially acute in low and middle-income countries (LMICs). Hence, any simple, cheap and flexible means of evaluating the incidence and evolution of the epidemic in a given country with a reasonable level of accuracy is useful. In this paper, we propose a technique based on (anonymous) surveys in which participants report on the health status of their contacts. This indirect reporting technique, known in the literature as network scale-up method, preserves the privacy of the participants and their contacts, and collects information from a larger fraction of the population (as compared to individual surveys). This technique has been deployed in the CoronaSurveys project, which has been collecting reports for the COVID-19 pandemic for more than two months. Results obtained by CoronaSurveys show the power and flexibility of the approach, suggesting that it could be an inexpensive and powerful tool for LMICs.

cs.DC↗

arXiv2020

Efficient graph construction for image representation

Graphs are useful to interpret widely used image processing methods, e.g., bilateral filtering, or to develop new ones, e.g., kernel based techniques. However, simple graph constructions are often used, where edge weight and connectivity depend on a few parameters. In particular, the sparsity of the graph is determined by the choice of a window size. As an alternative, we extend and adapt to images recently introduced non negative kernel regression (NNK) graph construction. In NNK graphs sparsity adapts to intrinsic data properties. Moreover, while previous work considered NNK graphs in generic settings, here we develop novel algorithms that take advantage of image properties so that the NNK approach can scale to large images. Our experiments show that sparse NNK graphs achieve improved energy compaction and denoising performance when compared to using graphs directly derived from the bilateral filter.

eess.IV↗

arXiv2020

Region adaptive graph fourier transform for 3d point clouds

We introduce the Region Adaptive Graph Fourier Transform (RA-GFT) for compression of 3D point cloud attributes. The RA-GFT is a multiresolution transform, formed by combining spatially localized block transforms. We assume the points are organized by a family of nested partitions represented by a rooted tree. At each resolution level, attributes are processed in clusters using block transforms. Each block transform produces a single approximation (DC) coefficient, and various detail (AC) coefficients. The DC coefficients are promoted up the tree to the next (lower resolution) level, where the process can be repeated until reaching the root. Since clusters may have a different numbers of points, each block transform must incorporate the relative importance of each coefficient. For this, we introduce the $\mathbf{Q}$-normalized graph Laplacian, and propose using its eigenvectors as the block transform. The RA-GFT achieves better complexity-performance trade-offs than previous approaches. In particular, it outperforms the Region Adaptive Haar Transform (RAHT) by up to 2.5 dB, with a small complexity overhead.

cs.CV↗

arXiv2020

An efficient algorithm for graph Laplacian optimization based on effective resistances

In graph signal processing, data samples are associated to vertices on a graph, while edge weights represent similarities between those samples. We propose a convex optimization problem to learn sparse well connected graphs from data. We prove that each edge weight in our solution is upper bounded by the inverse of the distance between data features of the corresponding nodes. We also show that the effective resistance distance between nodes is upper bounded by the distance between nodal data features. Thus, our proposed method learns a sparse well connected graph that encodes geometric properties of the data. We also propose a coordinate minimization algorithm that, at each iteration, updates an edge weight using exact minimization. The algorithm has a simple and low complexity implementation based on closed form expressions.

eess.SP↗

arXiv2020

Eigendecomposition-Free Sampling Set Selection for Graph Signals

This paper addresses the problem of selecting an optimal sampling set for signals on graphs. The proposed sampling set selection (SSS) is based on a localization operator that can consider both vertex domain and spectral domain localization. We clarify the relationships among the proposed method, sensor position selection methods in machine learning, and conventional SSS methods based on graph frequency. In contrast to the conventional graph signal processing-based approaches, the proposed method does not need to compute the eigendecomposition of a variation operator, while still considering (graph) frequency information. We evaluate the performance of our approach through comparisons of prediction errors and execution time.

eess.SP↗

arXiv2020

Graph Vertex Sampling with Arbitrary Graph Signal Hilbert Spaces

Graph vertex sampling set selection aims at selecting a set of ver-tices of a graph such that the space of graph signals that can be reconstructed exactly from those samples alone is maximal. In this context, we propose to extend sampling set selection based on spectral proxies to arbitrary Hilbert spaces of graph signals. Enabling arbitrary inner product of graph signals allows then to better account for vertex importance on the graph for a sampling adapted to the application. We first state how the change of inner product impacts sampling set selection and reconstruction, and then apply it in the context of geometric graphs to highlight how choosing an alternative inner product matrix can help sampling set selection and reconstruction.

eess.SP↗

arXiv2020

Perceptually inspired weighted MSE optimization using irregularity-aware graph Fourier transform

In image and video coding applications, distortion has been traditionally measured using mean square error (MSE), which suggests the use of orthogonal transforms, such as the discrete cosine transform (DCT). Perceptual metrics such as Structural Similarity (SSIM) are typically used after encoding, but not tied to the encoding process. In this paper, we consider an alternative framework where the goal is to optimize a weighted MSE metric, where different weights can be assigned to each pixel so as to reflect their relative importance in terms of perceptual image quality. For this purpose, we propose a novel transform coding scheme based on irregularity-aware graph Fourier transform (IAGFT), where the induced IAGFT is orthogonal, but the orthogonality is defined with respect to an inner product corresponding to the weighted MSE. We propose to use weights derived from local variances of the input image, such that the weighted MSE aligns with SSIM. In this way, the associated IAGFT can achieve a coding efficiency improvement in SSIM with respect to conventional transform coding based on DCT. Our experimental results show a compression gain in terms of multi-scale SSIM on test images.

eess.IV↗

arXiv2019

Reducing Storage in Large-Scale Photo Sharing Services using Recompression

The popularity of photo sharing services has increased dramatically in recent years. Increases in users, quantity of photos, and quality/resolution of photos combined with the user expectation that photos are reliably stored indefinitely creates a growing burden on the storage backend of these services. We identify a new opportunity for storage savings with application-specific compression for photo sharing services: photo recompression. We explore new photo storage management techniques that are fast so they do not adversely affect photo download latency, are complementary to existing distributed erasure coding techniques, can efficiently be converted to the standard JPEG user devices expect, and significantly increase compression. We implement our photo recompression techniques in two novel codecs, ROMP and L-ROMP. ROMP is a lossless JPEG recompression codec that compresses typical photos 15% over standard JPEG. L-ROMP is a lossy JPEG recompression codec that distorts photos in a perceptually un-noticeable way and typically achieves 28% compression over standard JPEG. We estimate the benefits of our approach on Facebook's photo stack and find that our approaches can reduce the photo storage by 0.3-0.9x the logical size of the stored photos, and offer additional, collateral benefits to the photo caching stack, including 5-11% fewer requests to the backend storage, 15-31% reduction in wide-area bandwidth, and 16% reduction in external bandwidth.

eess.IV↗