Raquel Dosil

Saliency from hierarchical adaptation through decorrelation and variance normalization

This paper presents a novel approach to visual saliency that relies on a contextually adapted representation produced through adaptive whitening of color and scale features. Unlike previous models, the proposal is grounded on the specific adaptation of the basis of low level features to the statistical structure of the image. Adaptation is achieved through decorrelation and contrast normalization in several steps in a hierarchical approach, in compliance with coarse features described in biological visual systems. Saliency is simply computed as the square of the vector norm in the resulting representation. The performance of the model is compared with several state-of-the-art approaches, in predicting human fixations using three different eye-tracking datasets. Referring this measure to the performance of human priority maps, the model proves to be the only one able to keep the same behavior through different datasets, showing free of biases. Moreover, it is able to predict a wide set of relevant psychophysical observations, to our knowledge, not reproduced together by any other model before.

Decorrelation and Distinctiveness Provide with Human-Like Saliency

In this work, we show the capability of a new model of saliency, of reproducing remarkable psychophysical results. The model presents low computational complexity compared to other models of the state of the art. It is based in biologically plausible mechanisms: the decorrelation and the distinctiveness of local responses. Decorrelation of scales is obtained from principal component analysis of multiscale low level features. Distinctiveness is measured through the Hotelling’s T2 statistic. The model is conceived to be used in a machine vision system, in which attention would contribute to enhance performance together with other visual functions. Experiments demonstrate the consistency with a wide variety of psychophysical phenomena, that are referenced in the visual attention modeling literature, with results that outperform other state of the art models.

Decomposition of three-dimensional medical images into visual patterns

In this paper, we present a method for the decomposition of a volumetric image into its most relevant visual patterns , which we define as features associated to local energy maxima of the image. The method involves the clustering of a set of predefined bandpass energy filters according to their ability to segregate the different features in the image, thus generating a set of composite-feature detectors tuned to the specific visual patterns present in the data. Clustering is based on a measure of statistical dependence between pairs of frequency features. We will illustrate the applicability of the method to the initialization of a three-dimensional geodesic active model.

Generalized ellipsoids and anisotropic filtering for segmentation improvement in 3D medical imaging

Abstract Deformable models have demonstrated to be very useful techniques for image segmentation. However, they present several weak points. Two of the main problems with deformable models are the following: (1) results are often dependent on the initial model location, and (2) the generation of image potentials is very sensitive to noise. Modeling and preprocessing methods presented in this paper contribute to solve these problems. We propose an initialization tool to obtain a good approximation to global shape and location of a given object into a 3D image. We also introduce a novel technique for corner preserving anisotropic diffusion filtering to improve contrast and corner measures. This is useful for both guiding initialization (global shape) and subsequent deformation for fine tuning (local shape).

Motion representation using composite energy features

This work tackles the segmentation of apparent-motion from a bottom-up perspective. When no information is available to build prior high-level models, the only alternative are bottom-up techniques. Hence, the whole segmentation process relies on the suitability of the low-level features selected to describe motion. A wide variety of low-level spatio-temporal features have been proposed so far. However, all of them suffer from diverse drawbacks. Here, we propose the use of composite energy features in bottom-up motion segmentation to solve several of these problems. Composite energy features are clusters of energy filters-pairs of band-pass filters in quadrature-each one sensitive to a different set of scale, orientation, direction of motion and speed. They are grouped in order to reconstruct independent motion patterns in a video sequence. A composite energy feature, this is, the response of one of these clusters of filters, can be built as a combination of the responses of the individual filters. Therefore, it inherits the desirable properties of energy filters but providing a more complete representation of motion patterns. In this paper, we will present our approach for integration of composite features based on the concept of Phase Congruence. We will show some results that illustrate the capabilities of this low-level motion representation and its usefulness in bottom-up motion segmentation and tracking.

Dissimilarity measures for visual pattern partitioning

We define a visual pattern as an image feature with frequency components in a range of bands that are aligned in phase. A technique to partition an image into its visual patterns involves clustering of the band-pass filtered versions of the image according to a measure of congruence in phase or, equivalently, alignment in the filters responses energy maxima. In this paper we study some measures of dissimilarity between images and discuss their suitability to the specific task of misalignment estimation between energy maps.

Integrating prior shape models into level-set approaches

To incorporate prior shape information into a deformable model either local or global shape modeling must be carried out. Local shape modeling involves manual interaction to accumulate information on the shape variability of any object. It depends on the existence of homologous points, or landmarks, that must be unambiguously and consistently located in different specimens. Global shape modeling does not require the existence of landmarks. Global properties can be characterized using only a few parameters, and tend to be much more stable than local properties.In this work we propose a new approach that combines the benefits of local and global shape modeling in the field of level-set approaches. The method starts with local shape parameterization, which eases user interaction. Then, the shape is converted into an implicit representation which exploits the stability and compactness of global shape parameters.

A new radial symmetry measure applied to photogrammetry

This work presents a new measure for radial symmetry and an algorithm for its computation. This measure identifies radially symmetric blobs as locations with contributions from all orientations at some scale. Hence, at a given scale, radial symmetry is computed as the product of the responses of a set of even symmetric feature detectors, with different orientations. This operator presents low sensitivity to shapes lacking radial symmetry, is robust to noise, contrast changes and strong perspective distortions, and shows a narrow point spread function. A multi-resolution measure is provided, computed as the maximum of the symmetry measure evaluated over a set of scales. We have applied this measure in the field of photogrammetry for the detection of circular coded fiducial targets. The detection of local maxima of multi-resolution radial symmetry is combined with a step of false-positive rejection, based on elliptical model fitting. In our experiments, the efficiency of target detection with this method is improved regarding a well-known commercial system, which is expected to improve the performance of bundle adjustment techniques. In order to fulfill all steps previous to bundle adjustment, we have also developed our own method for recognition of coded targets. This is accomplished by a standard procedure of segmentation and decoding of the ring sequence. Nevertheless, we have included a step for the verification of false positives of decoding based on correlation with reference targets. As far as we know, this approach cannot be found in literature.

Saliency Based on Decorrelation and Distinctiveness of Local Responses

In this paper we validate a new model of bottom-up saliency based in the decorrelation and the distinctiveness of local responses. The model is simple and light, and is based on biologically plausible mechanisms. Decorrelation is achieved by applying principal components analysis over a set of multiscale low level features. Distinctiveness is measured using the Hotellings T2 statistic. The presented approach provides a suitable framework for the incorporation of top-down processes like contextual priors, but also learning and recognition. We show its capability of reproducing human fixations on an open access image dataset and we compare it with other recently proposed models of the state of the art.

Scene Recognition through Visual Attention and Image Features: A Comparison between SIFT and SURF Approaches

In this work we study how we can use a novel model of spatial saliency (visual attention) combined with image features to significantly accelerate a scene recognition application and, at the same time, preserve recognition performance. To do so, we use a mobile robotlike application where scene recognition is carried out through the use of image features to characterize the different scenarios, and the Nearest Neighbor rule to carry out the classification. SIFT and SURF are two recent and competitive alternatives to image local featuring that we compare through extensive experimental work. Results from the experiments show that SIFT features perform significantly better than SURF features achieving important reductions in the size of the database of prototypes without significant losses in recognition performance, and thus, accelerating scene recognition. Also, from the experiments it is concluded that SURF features are less distinctive when using very large databases of interest points, as it occurs in the present case. Visual attention is the process by which the Human Visual System (HVS) is able to select from a given scene regions of interest that contain salient information, and thus, reduce the amount of information to be processed (Treisman, 1980; Koch, 1985). In the last decade, several computational models biologically motivated have been released to implement visual attention in image and video processing (Itti, 2000; Garcia-Diaz, 2008). Visual attention has also been used to improve object recognition and scene analysis (Bonaiuto, 2005; Walther, 2005). In this chapter, we study the utility of using a novel model of spatial saliency to improve a scene recognition application by reducing the amount of prototypes needed to carry out the classification task. The application is based on mobile robot-like video sequences taken in indoor facilities formed by several rooms and halls. The aim is to recognize the different scenarios in order to provide the mobile robot system with general location data. The visual attention approach is a novel model of bottom-up saliency that uses local phase information of the input data where the statistic information of second order is deleted to achieve a Retinoptical map of saliency. The proposed approach joints computational mechanisms of the two hypotheses largely accepted in early vision: first, the efficient coding