Wendy Aguilar

A robust Graph Transformation Matching for non-rigid registration

In this paper, we propose a simple and highly robust point-matching method named Graph Transformation Matching (GTM) relying on finding a consensus nearest-neighbour graph emerging from candidate matches. The method iteratively eliminates dubious matches in order to obtain the consensus graph. The proposed technique is compared against both the Softassign algorithm and a combination of RANSAC and epipolar constraint. Among these three techniques, GTM demonstrates to yield the best results in terms of elimination of outliers. The algorithm is shown to be able to deal with difficult cases such as duplication of patterns and non-rigid deformations of objects. An execution time comparison is also presented, where GTM shows to be also superior to RANSAC for high outlier rates. In order to improve the performance of GTM for lower outlier rates, we present an optimised version of the algorithm. Lastly, GTM is successfully applied in the context of constructing mosaics of retinal images, where feature points are extracted from properly segmented binary images. Similarly, the proposed method could be applied to a number of other important applications.

Constellations and the unsupervised learning of graphs

In this paper, we propose a novel method for the unsupervised clustering of graphs in the context of the constellation approach to object recognition. Such method is an EM central clustering algorithm which builds prototypical graphs on the basis of fast matching with graph transformations. Our experiments, both with random graphs and in realistic situations (visual localization), show that our prototypes improve the set median graphs and also the prototypes derived from our previous incremental method. We also discuss how the method scales with a growing number of images.

Graph-based methods for retinal mosaicing and vascular characterization

In this paper, we propose a highly robust point-matching method (Graph Transformation Matching - GTM) relying on finding the consensus graph emerging from putative matches. Such method is a two-phased one in the sense that after finding the consensus graph it tries to complete it as much as possible. We successfully apply GTM to image registration in the context of finding mosaics from retinal images. Feature points are obtained after properly segmenting such images. In addition, we also introduce a novel topological descriptor for quantifying disease by characterizing the arterial/venular trees. Such descriptor relies on diffusion kernels on graphs. Our experiments have showed only statistical significance for the case of arterial trees, which is consistent with previous findings.

The Past, Present, and Future of Artificial Life

For millennia people have wondered what makes the living different from the non-living. Beginning in the mid-1980s, artificial life has studied living systems using a synthetic approach: build life in order to understand it better, be it by means of software, hardware, or wetware. This review provides a summary of the advances that led to the development of artificial life, its current research topics, and open problems and opportunities. We classify artificial life research into fourteen themes: origins of life, autonomy, self-organization, adaptation (including evolution, development, and learning), ecology, artificial societies, behavior, computational biology, artificial chemistries, information, living technology, art, and philosophy. Being interdisciplinary, artificial life seems to be losing its boundaries and merging with other fields.

Contextual visual localization: cascaded submap classification, optimized saliency detection, and fast view matching

In this paper, we present a novel coarse-to-fine visual localization approach: contextual visual localization. This approach relies on three elements: (i) a minimal-complexity classifier for performing fast coarse localization (submap classification); (ii) an optimized saliency detector which exploits the visual statistics of the submap; and (iii) a fast view-matching algorithm which filters initial matchings with a structural criterion. The latter algorithm yields fine localization. Our experiments show that these elements have been successfully integrated for solving the global localization problem. Context, that is, the awareness of being in a particular submap, is defined by a supervised classifier tuned for a minimal set of features. Visual context is exploited both for tuning (optimizing) the saliency detection process, and to select potential matching views in the visual database, close enough to the query view.

Dev E-R

This paper describes a computational model named Dev E-R (Developmental Engagement-Reflection) that, inspired by Piagets theory, simulates the assimilation-accommodation adaptation process. It is implemented with a new extended version of the computational model of creativity known as Engagement-Reflection. That is, this model simulates adaptation as a creative activity. We introduce here the implementation of our model on an agent that is initialized with basic reflex conducts and that through the interaction with a 3D virtual world, it is able to build new behaviors autonomously. The new acquired skills, according to Piagets theory, are typically observed in children that have reached the second substage of the sensorimotor period.

Region and constellations based categorization of images with unsupervised graph learning

In this paper, we address the problem of image categorization with a fast novel method based on the unsupervised clustering of graphs in the context of both region-based segmentation and the constellation approach to object recognition. Such method is an EM central clustering algorithm which builds prototypical graphs on the basis of either Softassign or fast matching with graph transformations. We present two realistic applications and their experimental results: categorization of image segmentations and visual localization. We compare our graph prototypes with the set median graphs. Our results reveal that, on the one hand, structure extracted from images improves appearance-based visual localization accuracy. On the other hand, we show that the cost of our central graph clustering algorithm is the cost of a pairwise algorithm. We also discuss how the method scales with an increasing amount of images. In addition, we address the scientific question of what are the bounds of structural learning for categorization. Our in-depth experiments both for region-based and feature-based image categorization, will show that such bounds depend hardly on structural variability.

Vision-Based Localization for Mobile Robots Using a Set of Known Views

A robot localization scheme is presented in which a mobile robot finds its position within a known environment through image comparison. The images being compared are those taken by the robot throughout its reconnaissance trip and those stored in an image database that contains views taken from strategic positions within the environment, and that also contain position and orientation information. Image comparison is carried out using a scale-dependent keypoint-matching technique based on SIFT features, followed by a graph-based outlier elimination technique known as Graph Transformation Matching. Two techniques for position and orientation estimation are tested (epipolar geometry and clustering), followed by a probabilistic approach to position tracking (based on Monte Carlo localization).

Integrating Graph-Based Vision Perception to Spoken Conversation in Human-Robot Interaction

In this paper we present the integration of graph-based visual perception to spoken conversation in human-robot interaction. The proposed architecture has a dialogue manager as the central component for the multimodal interaction, which directs the robots behavior in terms of the intentions and actions associated to the conversational situations. We tested this ideas on a mobile robot programmed to act as a visitors guide to our department of computer science.

On the Selection of a Classification Technique for the Representation and Recognition of Dynamic Gestures

Previous evaluations of gesture recognition techniques have been focused on classification performance, while ignoring other relevant issues such as knowledge description, feature selection, error distribution and learning performance. In this paper, we present an empirical comparison of decision trees, neural networks and hidden Markov models for visual gesture recognition following these criteria. Our results show that none of these techniques is a definitive alternative for all these issues. While neural nets and hidden Markov models show the highest recognition rates, they sacrifice clarity of its knowledge; decision trees, on the other hand, are easy to create and analyze. Moreover, error dispersion is higher with neural nets. This information could be useful to develop a general computational theory of gestures. For the experiments, a database of 9 gestures with more than 7000 samples taken from 15 people was used.