Papers by Vittorio Murino
Real-time adaptive regulation of a visual camera for automatic investigation of changing environments
Proceedings of IECON '93 - 19th Annual Conference of IEEE Industrial Electronics
ABSTRACT
2008 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, 2008
3D face recognition(s) systems improve current 2D image-based approaches, but in general they are... more 3D face recognition(s) systems improve current 2D image-based approaches, but in general they are required to deal with larger amounts of data. Therefore, a compact representation of 3D faces is often crucial for a better manipulation of data, in the context of 3D face applications such as smart card identity verification systems. We propose a new compact 3D representation by focusing on the most significant parts of the face. We introduce a generative learning approach by adapting Hidden Markov Models (HMM) to work on 3D meshes. The geometry of local area around face fiducial points is modeled by training HMMs which provide a robust pose invariant point signature. Such description allows the matching by comparing the signature of corresponding points in a maximum-likelihood principle. We show that our descriptor is robust for recognizing expressions and performs faster than the current ICPbased 3D face recognition systems by maintaining a satisfactory recognition rate. Preliminary results on a subset of the FRGC 2.0 dataset are reported by considering subjects under different expressions.
Journal of Machine Learning Research, 2016
This paper presents a general vector-valued reproducing kernel Hilbert spaces (RKHS) framework fo... more This paper presents a general vector-valued reproducing kernel Hilbert spaces (RKHS) framework for the problem of learning an unknown functional dependency between a structured input space and a structured output space. Our formulation encompasses both Vector-valued Manifold Regularization and Co-regularized Multi-view Learning, providing in particular a unifying framework linking these two important learning approaches. In the case of the least square loss function, we provide a closed form solution, which is obtained by solving a system of linear equations. In the case of Support Vector Machine (SVM) classification, our formulation generalizes in particular both the binary Laplacian SVM to the multi-class, multi-view settings and the multi-class Simplex Cone SVM to the semisupervised, multi-view settings. The solution is obtained by solving a single quadratic optimization problem, as in standard SVM, via the Sequential Minimal Optimization (SMO) approach. Empirical results obtained on the task of object recognition, using several challenging data sets, demonstrate the competitiveness of our algorithms compared with other state-of-the-art methods.
Hybrid approach for accurate echo detection in the formation of acoustic images
'Challenges of Our Changing Global Environment'. Conference Proceedings. OCEANS '95 MTS/IEEE
ABSTRACT
2009 IEEE 12th International Conference on Computer Vision, 2009
Hybrid generative-discriminative techniques and, in particular, generative score-space classifica... more Hybrid generative-discriminative techniques and, in particular, generative score-space classification methods have proven to be valuable approaches in tackling difficult object or scene recognition problems. A generative model over the available data for each image class is first learned, providing a relatively comprehensive statistical representation. As a result, meaningful new image features at different levels of the model become available, encoding the degree of fitness of the data with respect to the model at different levels. Such features, defining a score space, are then fed into a discriminative classifier which can exploit the intrinsic data separability. In this paper, we present a generative score-space technique which encapsulates the uncertainty present in the generative learning phase usually disregarded by the state-of-the-art methods. In particular, we propose the use of variational free energy terms as feature vectors, so that the degree of fitness of the data and the uncertainty over the generative process are included explicitly in the data description. The proposed method is automatically superior to a pure generative classification, and we also experimentally illustrate it on a wide selection of generative models applied to challenging benchmarks in hard computer vision tasks such as scene, object and shape recognition. In several instances, the proposed approach beats the current state of the art in classification performance, while relying on computationally inexpensive models.
9th European Signal Processing Conference (EUSIPCO 1998), 1998
Perception. We describe a recognition framework to generate a virtual environment through CAD-bas... more Perception. We describe a recognition framework to generate a virtual environment through CAD-based vision techniques from optical data. Descriptions of objects of the environment in terms of aspect graphs, and suitable recognition strategies for them are compiled off-line. A relational graph of image features is obtained on-line by processing optical data, and matching occurs between such a graph, and descriptions of objects in the framed scene. Multiresolution techniques are used in order to adapt recognition strategies to the distance and relevance of objects within the field of view.
Proceedings. First International Symposium on 3D Data Processing Visualization and Transmission
We present a hierarchical and robust algorithm addressing the problem of filtering and segmentati... more We present a hierarchical and robust algorithm addressing the problem of filtering and segmentatio n of three-dimensional acoustic images. This algorithm is based on the tensor voting approach-a unified computational framework for the inference of multip le salient structures. Unlike most previous approaches , no models or prior information of the underwater environment, nor the intensity information of acoustic images is considered in this algorithm. Salient str uctures and outlier noisy points are directly clustered in two steps according to both the density and the structural information of input data. Our experimental trials show promising results, very robust despite the low computational complexity.
Multiple geometrical filters for 3D scale-invariant image representation of planar objects
[1992] Proceedings of the IEEE-SP International Symposium on Time-Frequency and Time-Scale Analysis
A low-level approach for 3-D invariant recognition of planar object from a centered monocular vie... more A low-level approach for 3-D invariant recognition of planar object from a centered monocular view is presented. The proposed scheme uses a distributed bank of filters that perform geometrical mapping on spatial image domain to generate a multidimensional (4-D) description. This image representation achieves invariance to 2-D rotation, scaling and perspective projection. Invariant classification and pose estimation, in terms of
This paper introduces a novel piecewise formulation for Photometric Stereo on static scenes explo... more This paper introduces a novel piecewise formulation for Photometric Stereo on static scenes exploiting a set of multiple view constraints on the surface. Such constraints will help to recover geometrical properties of the object and to eliminate the global bas-relief ambiguity. On the other hand, the proposed piecewise formulation will help to model more complex photometric properties of the surfaces using local lighting models. The experimental results show performance of the proposed method against 3D ground truth.
1996 8th European Signal Processing Conference (EUSIPCO 1996), 1996
In this paper the use of 3rd-order cumulants, i.e. triple correlations, is proposed for texture a... more In this paper the use of 3rd-order cumulants, i.e. triple correlations, is proposed for texture analysis. Properties of such features are derived, with particular attention to insensitivity to symmetrically distributed noises and statistical estimate stabilility. Experimental evaluation of 3rd-order cumulants as descriptive features for textures is carried out in comparison with autocorrelation-based approaches.
In this paper, a new segmentation approach for sets of 3D unorganized p oints is proposed. The me... more In this paper, a new segmentation approach for sets of 3D unorganized p oints is proposed. The method is based on a clustering procedure that separates the modes of a non-parametric m ultimodal density, following the mean-shift paradigm. The main idea consists in projecting the source 3D data into a set of independent sub-spaces, forming a joint multidimensional space. Each sub-space describes a geometric a spe t of the data set, such as the normals and principal curvatures, so as a dense region in a particular sub-spa ce indicates a set of 3D points sharing a similar value of that feature. A non-parametric clustering method is applied in this joint space by using a multidimensional kernel. This kernel smoothly takes into account for all the subspaces, moving towards high density regions in the joint space, separating them and providing “natural” clusters of 3D points. The algorithm can be implemented very easily and only few parameters need to be freely tun ed. Experiments are repo...
2014 22nd European Signal Processing Conference (EUSIPCO), 2014
This paper presents a novel approach for room reconstruction using unknown sound signals generate... more This paper presents a novel approach for room reconstruction using unknown sound signals generated in different locations of the environment. The approach is very general, that is fully uncalibrated, i.e. the locations of microphones, sound events and room reflectors are not known a priori. We show that, even if this problem implies a highly non-linear cost function, it is still possible to provide a solution close to the global minimum. Synthetic experiments show the proposed optimization framework can achieve reasonable results even in the presence of signal noise.
Wavelet-based Processing of EEG Data for Brain-Computer Interfaces
2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05) - Workshops
IEE Proceedings I Communications, Speech and Vision, 1993
In this paper, Bayesian networks of Markov random fields (BN-MRFs) are proposed as a technique fo... more In this paper, Bayesian networks of Markov random fields (BN-MRFs) are proposed as a technique for representing and applying apriori knowledge at different abstraction levels inside a distributed image processing framework. It is shown that this approach, thanks to the common probabilistic basis of the two techniques, is able to combine in a natural way causal inference properties at different abstraction levels as provided by Bayesian networks with optimisation criteria usually applied to find the best configuration for an MRF. Examples of two-level BN-MRFs are given, where each node uses a coupled Markov random field which has to solve a coupled restoration and segmentation problem. Experiments are concerned with expert-driven registered segmentation and tracking of regions from image sequences.
PLoS biology, 2018
Sleep science is entering a new era, thanks to new data-driven analysis approaches that, combined... more Sleep science is entering a new era, thanks to new data-driven analysis approaches that, combined with mouse gene-editing technologies, show a promise in functional genomics and translational research. However, the investigation of sleep is time consuming and not suitable for large-scale phenotypic datasets, mainly due to the need for subjective manual annotations of electrophysiological states. Moreover, the heterogeneous nature of sleep, with all its physiological aspects, is not fully accounted for by the current system of sleep stage classification. In this study, we present a new data-driven analysis approach offering a plethora of novel features for the characterization of sleep. This novel approach allowed for identifying several substages of sleep that were hidden to standard analysis. For each of these substages, we report an independent set of homeostatic responses following sleep deprivation. By using our new substages classification, we have identified novel differences ...
A Hidden Markov Model-based approach to
In this paper, we address the problem of dense 3D reconstruction from multiple view images subjec... more In this paper, we address the problem of dense 3D reconstruction from multiple view images subject to strong lighting variations. In this regard, a new piecewise framework is proposed to explicitly take into account the change of illumination across several wide-baseline images. Unlike multiview stereo and multi-view photometric stereo methods, this pipeline deals with wide-baseline images that are uncalibrated, in terms of both camera parameters and lighting conditions. Such a scenario is meant to avoid use of any specific imaging setup and provide a tool for normal users without any expertise. To the best of our knowledge, this paper presents the first work that deals with such unconstrained setting. We propose a coarse-to-fine approach, in which a coarse mesh is first created using a set of geometric constraints and, then, fine details are recovered by exploiting photometric properties of the scene. Augmenting the fine details on the coarse mesh is done via a final optimization step. Note that the method does not provide a generic solution for multi-view photometric stereo problem but it relaxes several common assumptions of this problem. The approach scales very well in size given its piecewise nature, dealing with large scale optimization and with severe missing data. Experiments on a benchmark dataset Robot data-set show the method performance against 3D ground truth.
Artificial Intelligence in Medicine, 2016
Objective: High-throughput technologies have generated an unprecedented amount of high-dimensiona... more Objective: High-throughput technologies have generated an unprecedented amount of high-dimensional gene expression data. Algorithmic approaches could be extremely useful to distill information and derive compact interpretable representations of the statistical patterns present in the data. This paper proposes a mining approach to extract an informative representation of gene expression profiles based on a generative model called the Counting Grid (CG). Method: Using the CG model, gene expression values are arranged on a discrete grid, learned in a way that "similar" co-expression patterns are arranged in close proximity, thus resulting in an intuitive visualization of the dataset. More than this, the model permits to identify the genes that distinguish between classes (e.g. different types of cancer). Finally, each sample can be characterized with a discriminative signature-extracted from the model-that can be effectively employed for classification. Results: A thorough evaluation on several gene expression datasets demonstrate the suitability of the proposed approach from a twofold perspective: numerically, we reached state-of-the-art classification accuracies on 5 datasets out of 7, and similar results when the approach is tested in a gene selection setting (with a stability always above 0.87); clinically, by confirming that many of the genes highlighted by the model as significant play also a key role for cancer biology. Conclusion: The proposed framework can be successfully exploited to meaningfully visualize the samples; detect medically relevant genes; properly classify samples.
Associative and symbolic algorithms for viewpoint-independent object recognition
Conference Proceedings 1991 IEEE International Conference on Systems, Man, and Cybernetics
ABSTRACT
Uploads
Papers by Vittorio Murino