Publications

Group detection in crowds will play a key role in future behavior analysis surveillance systems. In this work we build a new Structural SVM-based learning framework able to solve the group detection task by exploiting annotated video data to deduce a sociologically motivated distance measure founded on Hall's proxemics and Granger's causality. We improve over state-of-the-art results even in the most crowded test scenarios, while keeping the classification time affordable for quasi-real time applications. A new scoring scheme specifically designed for the group detection task is also proposed.

The analysis of anti-nuclear antibodies in HEp-2 cells by indirect immunofluorescence (IIF) is fundamental for the diagnosis of important immune pathologies; in particular, classifying the staining pattern of the cell is critical for the differential diagnosis of several types of diseases. Current tests based on human evaluation are time-consuming and suffer from very high variability, which impacts on the reliability of the results. As a solution to this problem, in this work we propose a technique that performs automated classification of the staining pattern. Our method combines textural feature extraction and a two-step feature selection scheme to select a limited number of image attributes that are best suited to the classification purpose and then recognizes the staining pattern by means of a Support Vector Machine module. Experiments on IIF images showed that our method is able to identify staining patterns with average accuracy of about 87%.

Many essential cell processes, such as the conformation of embedded proteins, membrane permeability, interaction with drugs and signalling, are directly connected to the molecular dynamics of cell membranes. The importance of this biology has led to an intensifying demand for hardware and software optimized models and tools, implemented on commodity high performance low-cost hardware, in order to provide the scientific community with virtual low cost laboratories. In the light of these considerations, we implemented an accelerated version of a molecular dynamics coarse-grain lipid bilayers simulator on commodity Graphic Processing Units (GPU) architectures. The characteristics of this molecular dynamics model, such as new force fields for pair potentials that include an unconventional representation for water and charges, were particularly challenging. We introduced new algorithms and data structures required by coarse-grain models compared to atomistic ones, for the modelling of the integration timestep, neighbour list generation, and nonbonded force interactions. We characterized the impact on performance of biological systems of differing complexity in terms of size, particle type and timestep. We also compared the simulations of many particle-type systems against single particle-type systems, to evaluate the overhead of additional structures needed to model more complex molecules. Moreover, we performed a detailed analysis on the profiling of the simulation code and its execution flows due to the computation of the non-bonded forces. Finally, we characterized the acceleration and accuracy of the simulations on three GPUs having different computation capabilities and parallelism, achieving one order of magnitude faster simulation execution times.

Analysis of 2D-GE images is a hot topic in bioinformatics research, since currently available commercial and academic software has proven to be not really effective and not completely automatic, often requiring manual revision of spots detection and refinement of computer generated matches. In this work, we present an effective technique for the detection and the reconstruction of over-saturated protein spots. Firstly, it reveals overexposed areas where spots may be truncated, and plateau regions caused by smeared and overlapped spots. As next, the correct distribution of pixel values in the overexposed areas and plateau regions is recovered by a two-dimensional fitting based on a generalized Gaussian distribution approximating the spots volume. Pixel correction according to the generalized Gaussian curve in saturated and smeared spots allows more accurate quantifications, providing more reliable image analysis results. As validation, we process highly exposed 2D-GE image, containing saturate spots, with respect to the corresponding non-saturated image, confirming that the method can effectively fix the saturated spots and enable correct spots quantification.

The new protocol for sequencing the messenger RNA in a cell, named RNA-seq produce millions of short sequence fragments. Next Generation Sequencing technology allows more accurate analysis but increase needs in term of computational resources. This paper describes the optimization of a RNA-seq analysis pipeline devoted to splicing variants detection, aimed at reducing computation time and providing a multi-user/multisample environment. This work brings two main contributions. First, we optimized a well-known algorithm called TopHat by parallelizing some sequential mapping steps. Second, we designed and implemented a hybrid virtual GRID infrastructure allowing to efficiently execute multiple instances of TopHat running on different samples or on behalf of different users, thus optimizing the overall execution time and enabling a flexible multi-user environment.

Chimeric phenomena have been recently recognized to play a significant role in the investigation and understanding of the fundamental mechanisms behind highly diffused pathologies such as tumors. In this paper we present a new methodology for the detection of fusion transcript from Next Generation Sequencing (NGS) data. The methodology exploits short paired-end reads coming from RNA-Seq experiments to determine a list of fused genes and to exactly identify the fusion boundaries, so that the exact chimeric sequence can be analysed. Both known and unknown transcripts are considered, enabling the detection of fusions involving unannotated genes. An automated toolflow that reports a set of candidate fused genes and the associated junctions has been implemented and applied to a publicly available data set of melanoma.

Color signatures, histograms and bag of colors are basic and effective strategies for describing the color content of images, for retrieving images by their color appearance or providing color annotation. In some domains, colors assume a specific meaning for users and the color-based classification and retrieval should mirror the initial suggestions given by users in the training set. For instance in fashion world, the names given to the dominant color of a garment or a dress reflect the fashion dictact and not an uniform division of the color space.In this paper we propose a general approach to implement color signature as a trained bag of words, defined on the basis of user defined color classes. The novel Class-based Color Bag of Words is a easy computable bag of words of color, constructed following an approach similar to the Median Cut algorithm, but biased by color distribution in the trained classes. Moreover, to dramatically reduce the computational effort we propose 3D integral histograms, a 3D extension of integral images, easily extensible for many histogram-based signature in 3D color space. Several comparisons in large fashion datasets confirm the discriminant power of this signature.

In the past literature, online alarm-based video-surveillance and offline forensic-based data mining systems are often treated separately, even from different scientific communities. However, the founding techniques are almost the same and, despite some examples in commercial systems, the cases on which an integrated approach is followed are limited. For this reason, this study describes an integrated tool capable of putting together these two subsystems in an effective way. Despite its generality, the proposal is here reported in the case of people trajectory analysis, both in real time and offline. Trajectories are modelled based on either their spatial location or their shape, and proper similarity measures are proposed. Special solutions to meet real-time requirements in both cases are also presented and the trade-off between efficiency and efficacy is analysed by comparing when using a statistical model and when not. Examples of results in large datasets acquired in the University campus are reported as preliminary evaluation of the system.

For quite some time, libraries, document and historical centers from opposite corners of the world have been the caretakers of our rich and assorted social legacy. They have protected and furnished access to the testimonies of knowledge, beauty and inspiration, such as sculptures, paintings, music and literature. The new information technologies have created unbelievable opportunities to make this common heritage more accessible for all. Culture is following the digital path and “memory institutions” are adapting the way in which they communicate with their public. Multimedia technologies have recently created the conditions for a true revolution in the cultural heritage area, with reference to the study, valorization, and fruition of artistic works. New multimedia technologies shall be able to be utilized to plan unique approaches to the perception and fulfillment of the masterful legacy, for instance, through smart cultural objects and new interfaces with the backing of items such as story-telling, gaming and learning.All the plurality of masterpieces (paintings, books, manuscripts, even photos of sculptures and architecture) can be effectively embedded into a unique ``paradigm'' through digitization. This allows a significant reduction in costs, an enormous expansion of public accessibility (and therefore income), and at the same time a tremendous freedom for data elaboration. In brief, digitization enhances pleasure for the public and usefulness to experts on cultural heritage assets.