Matteo Santoro

Solving structured sparsity regularization with proximal methods

Proximal methods have recently been shown to provide effective optimization procedures to solve the variational problems defining the l1 regularization algorithms. The goal of the paper is twofold. First we discuss how proximal methods can be applied to solve a large class of machine learning algorithms which can be seen as extensions of l1 regularization, namely structured sparsity regularization. For all these algorithms, it is possible to derive an optimization procedure which corresponds to an iterative projection algorithm. Second, we discuss the effect of a preconditioning of the optimization procedure achieved by adding a strictly convex functional to the objective function. Structured sparsity algorithms are usually based on minimizing a convex (not strictly convex) objective function and this might lead to undesired unstable behavior. We show that by perturbing the objective function by a small strictly convex term we often reduce substantially the number of required computations without affecting the prediction performance of the obtained solution.

The MAGIC-5 Project: medical applications on a GRID infrastructure connection

The MAGIC-5 Project aims at developing computer aided detection (CAD) software for medical applications on distributed databases by means of a GRID infrastructure connection. The use of automatic systems for analyzing medical images is of paramount importance in the screening programs, due to the huge amount of data to check. Examples are: mammographies for breast cancer detection, computed-tomography (CT) images for lung cancer analysis, and the positron emission tomography (PET) imaging for the early diagnosis of the Alzheimer disease. The need for acquiring and analyzing data stored in different locations requires a GRID approach of distributed computing system and associated data management. The GRID technologies allow remote image analysis and interactive online diagnosis, with a relevant reduction of the delays actually associated to the screening programs. From this point of view, the MAGIC-5 Collaboration can be seen as a group of distributed users sharing their resources for implementing different virtual organizations (VO), each one aiming at developing screening programs, tele-training, tele-diagnosis and epidemiologic studies for a particular pathology.

PADDLE: proximal algorithm for dual dictionaries learning

Recently, considerable research efforts have been devoted to the design of methods to learn from data overcomplete dictionaries for sparse coding. However, learned dictionaries require the solution of a sparse approximation problem for coding new data. In order to overcome this drawback, we propose an algorithm aimed at learning both a dictionary and its dual: a linear mapping directly performing the coding. Our algorithm is based on proximal methods and jointly minimizes the reconstruction error of the dictionary and the coding error of its dual; the sparsity of the representation is induced by an l1-based penalty on its coefficients. Experimental results show that the algorithm is capable of recovering the expected dictionaries. Furthermore, on a benchmark dataset the image features obtained from the dual matrix yield state-of-the-art classification performance while being much less computational intensive.

A Semi-automated Method for the Measurement of the Fetal Nuchal Translucency in Ultrasound Images

Nowadays the measurement of the nuchal translucency thickness is being used as part of routine ultrasound scanning during the end of the first trimester of pregnancy, for the screening of chromosomal defects, as trisomy 21. Currently, the measurement is being performed manually by physicians. The measurement can take a long time for being accomplished, needs to be performed by highly skilled operators, and is prone to errors. In this paper we present an algorithm that automatically detects the border of the nuchal translucency, once a region of interest has been manually identified. The algorithm is based on the minimisation of a cost function, and the optimisation is performed using the dynamic programming paradigm. The method we present overcomes several of the drawbacks present in the state of the art algorithms.

On the impact of learning hierarchical representations for visual recognition in robotics

Recent developments in learning sophisticated, hierarchical image representations have led to remarkable progress in the context of visual recognition. While these methods are becoming standard in modern computer vision systems, they are rarely adopted in robotics. The question arises of whether solutions, which have been primarily developed for image retrieval, can perform well in more dynamic and unstructured scenarios. In this paper we tackle this question performing an extensive evaluation of state of the art methods for visual recognition on a iCub robot. We consider the problem of classifying 15 different objects shown by a human demonstrator in a challenging Human-Robot Interaction scenario. The classification performance of hierarchical learning approaches are shown to outperform benchmark solutions based on local descriptors and template matching. Our results show that hierarchical learning systems are computationally efficient and can be used for real-time training and recognition of objects.

A connectionist architecture for view-independent grip-aperture computation.

This paper addresses the problem of extracting view-invariant visual features for the recognition of object-directed actions and introduces a computational model of how these visual features are processed in the brain. In particular, in the test-bed setting of reach-to-grasp actions, grip aperture is identified as a good candidate for inclusion into a parsimonious set of hand high-level features describing overall hand movement during reach-to-grasp actions. The computational model NeGOI (neural network architecture for measuring grip aperture in an observer-independent way) for extracting grip aperture in a view-independent fashion was developed on the basis of functional hypotheses about cortical areas that are involved in visual processing. An assumption built into NeGOI is that grip aperture can be measured from the superposition of a small number of prototypical hand shapes corresponding to predefined grip-aperture sizes. The key idea underlying the NeGOI model is to introduce view-independent units (VIP units) that are selective for prototypical hand shapes, and to integrate the output of VIP units in order to compute grip aperture. The distinguishing traits of the NEGOI architecture are discussed together with results of tests concerning its view-independence and grip-aperture recognition properties. The overall functional organization of NEGOI model is shown to be coherent with current functional models of the ventral visual stream, up to and including temporal area STS. Finally, the functional role of the NeGOI model is examined from the perspective of a biologically plausible architecture which provides a parsimonious set of high-level and view-independent visual features as input to mirror systems.

Is There Sparsity Beyond Additive Models

Abstract In this work we are interested in the problems of supervised learning and variable selection when the input-output dependence is described by a nonlinear function depending on a few variables. Our goal is to devise a sparse nonparametric model, avoiding linear or additive models. The key intuition is to measure the importance of each variable in the model by making use of partial derivatives. Based on this idea we propose and study a new regularizer and a corresponding least squares regularization scheme. Using concepts and results from the theory of reproducing kernel Hilbert spaces and proximal methods, we show that the proposed learning algorithm induces a minimization problem which can be provably solved by an iterative procedure. The consistency properties of the obtained estimator are studied both in terms of prediction and selection performance.

Learning adaptive and sparse representations of medical images

In this paper we discuss the impact of using algorithms for dictionary learning to build adaptive and sparse representations of medical images. The effectiveness of coding data as sparse linear combinations of the elements of an over-complete dictionary is well assessed in the medical context. Confirming what has been observed for natural images, we show the benefits of using adaptive dictionaries, directly learned from a set of training images, that better capture the distribution of the data. The experiments focus on the specific task of image denoising and produce clear evidence of the benefits obtained with the proposed approach.

The Channeler Ant Model: Object segmentation with virtual ant colonies

3-D object segmentation is an important and challenging topic in computer vision that could be tackled with artificial life models. A Channeler Ant Model (CAM), based on the natural ant capabilities of dealing with 3-D environments through self-organization and emergent behaviours, is proposed. Ant colonies, defined in terms of moving, pheromone laying, reproduction, death and deviating behaviours rules, is able to segment artificially generated objects of different shape, intensity, background. The model depends on few parameters and provides an elegant solution for the segmentation of 3-D structures in noisy environments with unknown range of image intensities: even when there is a partial overlap between the intensity and noise range, it provides a complete segmentation with negligible contamination (i.e., fraction of segmented voxels that do not belong to the object).

A Database of Segmented MRI Images of the Wrist and the Hand in Patients with Rheumatic Diseases

This paper is concerned with the ideation, organization and distribution of a database of segmented MRI images - and associated clinical parameters - of the wrist and the hand in patients affected by a variety of the most frequent rheumatic diseases. The final goal is empowering future biomedical research thanks to the completeness of details and cases. MRI Images were analyzed by means of the software RheumaSCORE (Softeco Sismat Srl), which performs semi-automatic segmentation of the bones, returns the volume of bones and erosions, as well as their tri-dimensional reconstruction. In order to favor its exploitation, the database of segmented images, along with many relevant clinical anthropometric parameters, are available online through the Patient Browser platform (Softeco Sismat Srl). Moreover, the original images and their clinical parameters are accessible online through the dedicated DICOM viewer QuantaView (CAMELOT Biomedical Systems Srl).