A experiência como fator determinante na representação espacial do deficiente visual

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Deep learning architecture for gesture recognition

Pós-graduação em Engenharia Mecânica - FEG

Interface para classificação de dados por máquinas de vetores de suporte

In this project the Pattern Recognition Problem is approached with the Support Vector Machines (SVM) technique, a binary method of classification that provides the best solution separating the data in the better way with a hiperplan and an extension of the input space dimension, as a Machine Learning solution. The system aims to classify two classes of pixels chosen by the user in the interface in the interest selection phase and in the background selection phase, generating all the data to be used in the LibSVM library, a library that implements the SVM, illustrating the library operation in a casual way. The data provided by the interface is organized in three types, RGB (Red, Green and Blue color system), texture (calculated) or RGB + texture. At last the project showed successful results, where the classification of the image pixels was showed as been from one of the two classes, from the interest selection area or from the background selection area. The simplest user view of results classification is the RGB type of data arrange, because it’s the most concrete way of data acquisition

Color texture analysis based on fractal descriptors

Color texture classification is an important step in image segmentation and recognition. The color information is especially important in textures of natural scenes, such as leaves surfaces, terrains models, etc. In this paper, we propose a novel approach based on the fractal dimension for color texture analysis. The proposed approach investigates the complexity in R, G and B color channels to characterize a texture sample. We also propose to study all channels in combination, taking into consideration the correlations between them. Both these approaches use the volumetric version of the Bouligand-Minkowski Fractal Dimension method. The results show a advantage of the proposed method over other color texture analysis methods. (C) 2011 Elsevier Ltd. All rights reserved.

Fractal descriptors based on Fourier spectrum applied to texture analysis

This work proposes the development and study of a novel technique lot the generation of fractal descriptors used in texture analysis. The novel descriptors are obtained from a multiscale transform applied to the Fourier technique of fractal dimension calculus. The power spectrum of the Fourier transform of the image is plotted against the frequency in a log-log scale and a multiscale transform is applied to this curve. The obtained values are taken as the fractal descriptors of the image. The validation of the proposal is performed by the use of the descriptors for the classification of a dataset of texture images whose real classes are previously known. The classification precision is compared to other fractal descriptors known in the literature. The results confirm the efficiency of the proposed method. (C) 2012 Elsevier B.V. All rights reserved.

Risk factors for computer visual syndrome (CVS) among operators of two call centers in Sao Paulo, Brazil

The aims of this study were to investigate work conditions, to estimate the prevalence and to describe risk factors associated with Computer Vision Syndrome among two call centers' operators in Sao Paulo (n = 476). The methods include a quantitative cross-sectional observational study and an ergonomic work analysis, using work observation, interviews and questionnaires. The case definition was the presence of one or more specific ocular symptoms answered as always, often or sometimes. The multiple logistic regression model, were created using the stepwise forward likelihood method and remained the variables with levels below 5% (p < 0.05). The operators were mainly female and young (from 15 to 24 years old). The call center was opened 24 hours and the operators weekly hours were 36 hours with break time from 21 to 35 minutes per day. The symptoms reported were eye fatigue (73.9%), "weight" in the eyes (68.2%), "burning" eyes (54.6%), tearing (43.9%) and weakening of vision (43.5%). The prevalence of Computer Vision Syndrome was 54.6%. Associations verified were: being female (OR 2.6, 95% CI 1.6 to 4.1), lack of recognition at work (OR 1.4, 95% CI 1.1 to 1.8), organization of work in call center (OR 1.4, 95% CI 1.1 to 1.7) and high demand at work (OR 1.1, 95% CI 1.0 to 1.3). The organization and psychosocial factors at work should be included in prevention programs of visual syndrome among call centers' operators.

EDGE-PRESERVING SPECKLE TEXTURE REMOVAL BY INTERFERENCE-BASED SPECKLE FILTERING FOLLOWED BY ANISOTROPIC DIFFUSION

Ultrasonography has an inherent noise pattern, called speckle, which is known to hamper object recognition for both humans and computers. Speckle noise is produced by the mutual interference of a set of scattered wavefronts. Depending on the phase of the wavefronts, the interference may be constructive or destructive, which results in brighter or darker pixels, respectively. We propose a filter that minimizes noise fluctuation while simultaneously preserving local gray level information. It is based on steps to attenuate the destructive and constructive interference present in ultrasound images. This filter, called interference-based speckle filter followed by anisotropic diffusion (ISFAD), was developed to remove speckle texture from B-mode ultrasound images, while preserving the edges and the gray level of the region. The ISFAD performance was compared with 10 other filters. The evaluation was based on their application to images simulated by Field II (developed by Jensen et al.) and the proposed filter presented the greatest structural similarity, 0.95. Functional improvement of the segmentation task was also measured, comparing rates of true positive, false positive and accuracy. Using three different segmentation techniques, ISFAD also presented the best accuracy rate (greater than 90% for structures with well-defined borders). (E-mail: fernando.okara@gmail.com) (C) 2012 World Federation for Ultrasound in Medicine & Biology.

Texture analysis by multi-resolution fractal descriptors

This work proposes a novel texture descriptor based on fractal theory. The method is based on the Bouligand- Minkowski descriptors. We decompose the original image recursively into four equal parts. In each recursion step, we estimate the average and the deviation of the Bouligand-Minkowski descriptors computed over each part. Thus, we extract entropy features from both average and deviation. The proposed descriptors are provided by concatenating such measures. The method is tested in a classification experiment under well known datasets, that is, Brodatz and Vistex. The results demonstrate that the novel technique achieves better results than classical and state-of-the-art texture descriptors, such as Local Binary Patterns, Gabor-wavelets and co-occurrence matrix.

Combining fractal and deterministic walkers for texture analysis and classification

In this paper,we present a novel texture analysis method based on deterministic partially self-avoiding walks and fractal dimension theory. After finding the attractors of the image (set of pixels) using deterministic partially self-avoiding walks, they are dilated in direction to the whole image by adding pixels according to their relevance. The relevance of each pixel is calculated as the shortest path between the pixel and the pixels that belongs to the attractors. The proposed texture analysis method is demonstrated to outperform popular and state-of-the-art methods (e.g. Fourier descriptors, occurrence matrix, Gabor filter and local binary patterns) as well as deterministic tourist walk method and recent fractal methods using well-known texture image datasets.

Methodologies for visual correspondence

Visual correspondence is a key computer vision task that aims at identifying projections of the same 3D point into images taken either from different viewpoints or at different time instances. This task has been the subject of intense research activities in the last years in scenarios such as object recognition, motion detection, stereo vision, pattern matching, image registration. The approaches proposed in literature typically aim at improving the state of the art by increasing the reliability, the accuracy or the computational efficiency of visual correspondence algorithms. The research work carried out during the Ph.D. course and presented in this dissertation deals with three specific visual correspondence problems: fast pattern matching, stereo correspondence and robust image matching. The dissertation presents original contributions to the theory of visual correspondence, as well as applications dealing with 3D reconstruction and multi-view video surveillance.

Neural basis of visual deficits recovery: visual residual functions and multisensory integration.

The ability of integrating into a unified percept sensory inputs deriving from different sensory modalities, but related to the same external event, is called multisensory integration and might represent an efficient mechanism of sensory compensation when a sensory modality is damaged by a cortical lesion. This hypothesis has been discussed in the present dissertation. Experiment 1 explored the role of superior colliculus (SC) in multisensory integration, testing patients with collicular lesions, patients with subcortical lesions not involving the SC and healthy control subjects in a multisensory task. The results revealed that patients with collicular lesions, paralleling the evidence of animal studies, demonstrated a loss of multisensory enhancement, in contrast with control subjects, providing the first lesional evidence in humans of the essential role of SC in mediating audio-visual integration. Experiment 2 investigated the role of cortex in mediating multisensory integrative effects, inducing virtual lesions by inhibitory theta-burst stimulation on temporo-parietal cortex, occipital cortex and posterior parietal cortex, demonstrating that only temporo-parietal cortex was causally involved in modulating the integration of audio-visual stimuli at the same spatial location. Given the involvement of the retino-colliculo-extrastriate pathway in mediating audio-visual integration, the functional sparing of this circuit in hemianopic patients is extremely relevant in the perspective of a multisensory-based approach to the recovery of unisensory defects. Experiment 3 demonstrated the spared functional activity of this circuit in a group of hemianopic patients, revealing the presence of implicit recognition of the fearful content of unseen visual stimuli (i.e. affective blindsight), an ability mediated by the retino-colliculo-extrastriate pathway and its connections with amygdala. Finally, Experiment 4 provided evidence that a systematic audio-visual stimulation is effective in inducing long-lasting clinical improvements in patients with visual field defect and revealed that the activity of the spared retino-colliculo-extrastriate pathway is responsible of the observed clinical amelioration, as suggested by the greater improvement observed in patients with cortical lesions limited to the occipital cortex, compared to patients with lesions extending to other cortical areas, found in tasks high demanding in terms of spatial orienting. Overall, the present results indicated that multisensory integration is mediated by the retino-colliculo-extrastriate pathway and that a systematic audio-visual stimulation, activating this spared neural circuit, is able to affect orientation towards the blind field in hemianopic patients and, therefore, might constitute an effective and innovative approach for the rehabilitation of unisensory visual impairments.

On-line adaptive visual tracking

Visual tracking is the problem of estimating some variables related to a target given a video sequence depicting the target. Visual tracking is key to the automation of many tasks, such as visual surveillance, robot or vehicle autonomous navigation, automatic video indexing in multimedia databases. Despite many years of research, long term tracking in real world scenarios for generic targets is still unaccomplished. The main contribution of this thesis is the definition of effective algorithms that can foster a general solution to visual tracking by letting the tracker adapt to mutating working conditions. In particular, we propose to adapt two crucial components of visual trackers: the transition model and the appearance model. The less general but widespread case of tracking from a static camera is also considered and a novel change detection algorithm robust to sudden illumination changes is proposed. Based on this, a principled adaptive framework to model the interaction between Bayesian change detection and recursive Bayesian trackers is introduced. Finally, the problem of automatic tracker initialization is considered. In particular, a novel solution for categorization of 3D data is presented. The novel category recognition algorithm is based on a novel 3D descriptors that is shown to achieve state of the art performances in several applications of surface matching.

Reliability and consistency affect the integration of visual depth cues

Abstract Originalsprache (englisch) Visual perception relies on a two-dimensional projection of the viewed scene on the retinas of both eyes. Thus, visual depth has to be reconstructed from a number of different cues that are subsequently integrated to obtain robust depth percepts. Existing models of sensory integration are mainly based on the reliabilities of individual cues and disregard potential cue interactions. In the current study, an extended Bayesian model is proposed that takes into account both cue reliability and consistency. Four experiments were carried out to test this model's predictions. Observers had to judge visual displays of hemi-cylinders with an elliptical cross section, which were constructed to allow for an orthogonal variation of several competing depth cues. In Experiment 1 and 2, observers estimated the cylinder's depth as defined by shading, texture, and motion gradients. The degree of consistency among these cues was systematically varied. It turned out that the extended Bayesian model provided a better fit to the empirical data compared to the traditional model which disregards covariations among cues. To circumvent the potentially problematic assessment of single-cue reliabilities, Experiment 3 used a multiple-observation task, which allowed for estimating perceptual weights from multiple-cue stimuli. Using the same multiple-observation task, the integration of stereoscopic disparity, shading, and texture gradients was examined in Experiment 4. It turned out that less reliable cues were downweighted in the combined percept. Moreover, a specific influence of cue consistency was revealed. Shading and disparity seemed to be processed interactively while other cue combinations could be well described by additive integration rules. These results suggest that cue combination in visual depth perception is highly flexible and depends on single-cue properties as well as on interrelations among cues. The extension of the traditional cue combination model is defended in terms of the necessity for robust perception in ecologically valid environments and the current findings are discussed in the light of emerging computational theories and neuroscientific approaches.

Residual visual processing following real or virtual lesions to primary visual pathways

Lesions to the primary geniculo-striate visual pathway cause blindness in the contralesional visual field. Nevertheless, previous studies have suggested that patients with visual field defects may still be able to implicitly process the affective valence of unseen emotional stimuli (affective blindsight) through alternative visual pathways bypassing the striate cortex. These alternative pathways may also allow exploitation of multisensory (audio-visual) integration mechanisms, such that auditory stimulation can enhance visual detection of stimuli which would otherwise be undetected when presented alone (crossmodal blindsight). The present dissertation investigated implicit emotional processing and multisensory integration when conscious visual processing is prevented by real or virtual lesions to the geniculo-striate pathway, in order to further clarify both the nature of these residual processes and the functional aspects of the underlying neural pathways. The present experimental evidence demonstrates that alternative subcortical visual pathways allow implicit processing of the emotional content of facial expressions in the absence of cortical processing. However, this residual ability is limited to fearful expressions. This finding suggests the existence of a subcortical system specialised in detecting danger signals based on coarse visual cues, therefore allowing the early recruitment of flight-or-fight behavioural responses even before conscious and detailed recognition of potential threats can take place. Moreover, the present dissertation extends the knowledge about crossmodal blindsight phenomena by showing that, unlike with visual detection, sound cannot crossmodally enhance visual orientation discrimination in the absence of functional striate cortex. This finding demonstrates, on the one hand, that the striate cortex plays a causative role in crossmodally enhancing visual orientation sensitivity and, on the other hand, that subcortical visual pathways bypassing the striate cortex, despite affording audio-visual integration processes leading to the improvement of simple visual abilities such as detection, cannot mediate multisensory enhancement of more complex visual functions, such as orientation discrimination.

Classificazione di oggetti in immagini attraverso il modello Bag of Visual Words

Generic object recognition is an important function of the human visual system and everybody finds it highly useful in their everyday life. For an artificial vision system it is a really hard, complex and challenging task because instances of the same object category can generate very different images, depending of different variables such as illumination conditions, the pose of an object, the viewpoint of the camera, partial occlusions, and unrelated background clutter. The purpose of this thesis is to develop a system that is able to classify objects in 2D images based on the context, and identify to which category the object belongs to. Given an image, the system can classify it and decide the correct categorie of the object. Furthermore the objective of this thesis is also to test the performance and the precision of different supervised Machine Learning algorithms in this specific task of object image categorization. Through different experiments the implemented application reveals good categorization performances despite the difficulty of the problem. However this project is open to future improvement; it is possible to implement new algorithms that has not been invented yet or using other techniques to extract features to make the system more reliable. This application can be installed inside an embedded system and after trained (performed outside the system), so it can become able to classify objects in a real-time. The information given from a 3D stereocamera, developed inside the department of Computer Engineering of the University of Bologna, can be used to improve the accuracy of the classification task. The idea is to segment a single object in a scene using the depth given from a stereocamera and in this way make the classification more accurate.