SciTePress - Affiliation Profile

University of Verona, Italy

Context (please use complete words only)

Field

Sort by

7 Papers

A Novel Dictionary Learning based Multiple Instance Learning Approach to Action Recognition from Videos

Topics: Multi-Instance Learning; Instance-Based Learning; Learning of Action Patterns

Authors: Abhinaba Roy ¹ , Biplab Banerjee ² and Vittorio Murino ¹

Affiliations: ¹ Istituto Italiano di Tecnologia, Italy , ² Indian Institute of Technology, India

Digital Object Identifier: 10.5220/0006200205190526

Proceedings of ICPRAM 2017, ISBN: 978-989-758-222-6

Quick Abstract

In this paper we deal with the problem of action recognition from unconstrained videos under the notion of multiple instance learning (MIL). The traditional MIL paradigm considers the data items as bags of instances with the constraint that the positive bags contain some class-specific instances whereas the negative bags consist of instances only from negative classes. A classifier is then further constructed using the bag level annotations and a distance metric between the bags. However, such an approach is not robust to outliers and is time consuming for a moderately large dataset. In contrast, we propose a dictionary learning based strategy to MIL which first identifies class-specific discriminative codewords, and then projects the bag-level instances into a probabilistic embedding space with respect to the selected codewords. This essentially generates a fixed-length vector representation of the bags which is specifically dominated by the properties of the class-specific instances. We introduce a novel exhaustive search strategy using a support vector machine classifier in order to highlight the class-specific codewords. The standard multiclass classification pipeline is followed henceforth in the new embedded feature space for the sake of action recognition. We validate the proposed framework on the challenging KTH and Weizmann datasets, and the results obtained are promising and comparable to representative techniques from the literature.

Related Ontology Subjects/Areas/Topics (7)

Applications	Multi-Instance Learning	Software Engineering
Instance-Based Learning	Pattern Recognition	Theory and Methods
Learning of Action Patterns

Segmentation of Retinal Ganglion Cells From Fluorescent Microscopy Imaging

Topics: Fluorescent Microscopy; Confocal Microscopy

Authors: Silvia Baglietto ^{1

;

2} , Ibolya E. Kepiro ¹ , Gerrit Hilgen ³ , Evelyne Sernagor ³ , Vittorio Murino ^{1

;

4} and Diego Sona ^{1

;

5}

Affiliations: ¹ Istituto Italiano di Tecnologia, Italy , ² University of Genova, Italy , ³ Institute of Neuroscience, United Kingdom , ⁴ University of Verona, Italy , ⁵ Fondazione Bruno Kessler, Italy

Digital Object Identifier: 10.5220/0006110300170023

Proceedings of BIOIMAGING 2017/BIOSTEC, ISBN: 978-989-758-215-8

Quick Abstract

The visual information processing starts in the retina. The working mechanisms of its complex stratified circuits, in which ganglion cells play a central role, is still largely unknown. Understanding the visual coding is a challenging and active research area also requiring automated analysis of retinal images. It demands appropriate algorithms and methods for studying a network population of strictly entangled cells. Within this framework, we propose a combined technique for segmenting retinal ganglion cell (RGC) bodies, the output elements of the retina. The method incorporates a blob enhancement filtering in order to select the specific cell shapes, an active contour process for precise border segmentation and a watershed transform step which separates single cell contours in possible grouped segmentations. The proposed approach has been validated on fluorescent microscopy images of mouse retinas with promising results.

GENERATIVE EMBEDDINGS BASED ON RICIAN MIXTURES - Application to Kernel-based Discriminative Classification of Magnetic Resonance Images

Topics: Kernel Methods; Classification; Medical Imaging

Authors: Anna C. Carli ¹ , Mario A. T. Figueiredo ² , Manuele Bicego ³ and Vittorio Murino ³

Affiliations: ¹ Università di Verona, Italy , ² IST-Instituto Superior Técnico e IT-Instituto de Telecomunicações, Portugal , ³ Università di Verona and Istituto Italiano di Tecnologia (IIT), Italy

Proceedings of ICPRAM 2012, ISBN: 978-989-8425-98-0

Quick Abstract

Most approaches to classifier learning for structured objects (such as images or sequences) are based on probabilistic generative models. On the other hand, state-of-the-art classifiers for vectorial data are learned discriminatively. In recent years, these two dual paradigms have been combined via the use of generative embeddings (of which the Fisher kernel is arguably the best known example); these embeddings are mappings from the object space into a fixed dimensional score space, induced by a generative model learned from data, on which a (maybe kernel-based) discriminative approach can then be used. This paper proposes a new semi-parametric approach to build generative embeddings for classification of magnetic resonance images (MRI). Based on the fact that MRI data is well described by Rice distributions, we propose to use Rician mixtures as the underlying generative model, based on which several different generative embeddings are built. These embeddings yield vectorial representations on which kernel-based support vector machines (SVM) can be trained for classification. Concerning the choice of kernel, we adopt the recently proposed nonextensive information theoretic kernels. The methodology proposed was tested on a challenging classification task, which consists in classifying MRI images as belonging to schizophrenic or non-schizophrenic human subjects. The classification is based on a set of regions of interest (ROIs) in each image, with the classifiers corresponding to each ROI being combined via boosting. The experimental results show that the proposed methodology outperforms the previous state-of-the-art methods on the same dataset.

Related Ontology Subjects/Areas/Topics (7)

Applications	Medical Imaging	Software Engineering
Classification	Pattern Recognition	Theory and Methods
Kernel Methods

AN INNOVATIVE PROTOCOL FOR COMPARING PROTEIN BINDING SITES VIA ATOMIC GRID MAPS

Topics: Bioinformatics & Pattern Discovery; Clustering and Classification Methods

Authors: M. Bicego ¹ , A. D. Favia ² , P. Bisignano ² , A. Cavalli ² and V. Murino ¹

Affiliations: ¹ Istituto Italiano di Tecnologia and University of Verona, Italy , ² Istituto Italiano di Tecnologia, Italy

Digital Object Identifier: 10.5220/0003637404130422

Proceedings of KDIR 2011/IC3K, ISBN: 978-989-8425-79-9

Quick Abstract

This paper deals with a novel computational approach that aims to measure the similarities of protein binding sites through comparison of atomic grid maps. The assessment of structural similarity between proteins is a longstanding goal in biology and in structure-based drug design. Instead of focusing on standard structural alignment techniques, mostly based on superposition of common structural elements, the proposed approach starts from a physicochemical description of the proteins’ binding site. We call these atomic grid maps. These maps are preprocessed to reduce the dimensionality of the data while retaining the relevant information. Then, we devise an alignment-based similarity measure, based on a rigid registration algorithm (the Iterative Closest Point –ICP). The proposed approach, tested on a real dataset involving 22 proteins, has shown encouraging results in comparison with standard procedures.

Related Ontology Subjects/Areas/Topics (6)

Artificial Intelligence		Clustering and Classification Methods		Knowledge-Based Systems
BioInformatics & Pattern Discovery		Knowledge Discovery and Information Retrieval		Symbolic Systems

BACKGROUND SUBTRACTION WITH ADAPTIVE SPATIO-TEMPORAL NEIGHBORHOOD ANALYSIS

Topic: Motion and Tracking

Authors: Marco Cristani and Vittorio Murino

Affiliation: Università degli Studi of Verona, Italy

Digital Object Identifier: 10.5220/0001072704840489

Proceedings of VISAPP 2008/VISIGRAPP, ISBN: 978-989-8111-21-0

Quick Abstract

In the literature, visual surveillance methods based on joint pixel and region analysis for background subtraction are proven to be effective in discovering foreground objects in cluttered scenes. Typically, per-pixel foreground detection is contextualized in a local neighborhood region in order to limit false alarms. However, such methods have an heavy computational cost, depending on the size of the surrounding region considered for each pixel. In this paper, we propose an original and efficient joint pixel-region analysis technique able to automatically select the sampling rate with which pixels in different areas are checked out, while adapting the size of the neighborhood region considered. The algorithm has been validated on standard videos with benchmark tests, proving the goodness of the approach, especially in terms of quality of the detection with respect to the frame rate achieved.

Related Ontology Subjects/Areas/Topics (8)

Applications	Methodologies and Methods	Pattern Recognition
Computer Vision, Visualization and Computer Graphics	Motion and Tracking	Physiological Computing Systems
Human-Computer Interaction	Motion, Tracking and Stereo Vision

Geo-located Image Categorization and Location Recognition

Authors: Marco Cristani , Alessandro Perina , Umberto Castellani and Vittorio Murino

Affiliation: Università degli Studi di Verona, Italy

Digital Object Identifier: 10.5220/0002340400930102

Proceedings of IMTA 2008/VISIGRAPP, ISBN: 978-989-8111-25-8

Quick Abstract

Image categorization is undoubtedly one of the most recent and challenging problems faced in Computer Vision. The scientific literature is plenty of methods more or less efficient and dedicated to a specific class of images; further, commercial systems are also going to be advertised in the market. Nowadays, additional data can also be attached to the images, enriching its semantic interpretation beyond the pure appearance. This is the case of geo-location data that contain information about the geographical place where an image has been acquired. This data allow, if not require, a different management of the images, for instance, to the purpose of easy retrieval from a repository, or of identifying the geographical place of an unknown picture, given a geo-referenced image repository. This paper constitutes a first step in this sense, presenting a method for geo-referenced image categorization, and for the recognition of the geographical location of an image without such information available. The solutions presented are based on robust pattern recognition techniques, such as the probabilistic Latent Semantic Analysis, the Mean Shift clustering and the Support Vector Machines. Experiments have been carried out on a couple of geographical image databases: results are actually very promising, opening new interesting challenges and applications in this research field.

A SPATIAL SAMPLING MECHANISM FOR EFFECTIVE BACKGROUND SUBTRACTION

Topic: Tracking of People and Surveillance

Authors: Marco Cristani and Vittorio Murino

Affiliation: Università degli Studi of Verona, Italy

Digital Object Identifier: 10.5220/0002058304030410

Proceedings of VISAPP 2007/VISIGRAPP, ISBN: 978-972-8865-74-0

Quick Abstract

In the video surveillance literature, background (BG) subtraction is an important and fundamental issue. In this context, a consistent group of methods operates at region level, evaluating in fixed zones of interest pixel values’ statistics, so that a per-pixel foreground (FG) labeling can be performed. In this paper, we propose a novel hybrid, pixel/region, approach for background subtraction. The method, named Spatial-Time Adaptive Per Pixel Mixture Of Gaussian (S-TAPPMOG), evaluates pixel statistics considering zones of interest that change continuously over time, adopting a sampling mechanism. In this way, numerous classical BG issues can be efficiently faced: actually, it is possible to model the background information more accurately in the chromatic uniform regions exhibiting stable behavior, thus minimizing foreground camouflages. At the same time, it is possible to model successfully regions of similar color but corrupted by heavy noise, in order to minimize false FG detections. Such approach, outperforming state of the art methods, is able to run in quasi-real time and it can be used at a basis for more structured background subtraction algorithms.

Related Ontology Subjects/Areas/Topics (3)

Computer Vision, Visualization and Computer Graphics

Motion, Tracking and Stereo Vision

Tracking of People and Surveillance

Topic(s):

Title:
ISBN:
Year:
Acronym:
Subject:

Title:
Author:
Affiliation:
Subject:

Name:
Affiliation:
Country:
Conference:
Subject:

Name:
Country:
Subject:

Affiliation

Filter listed results, by