Place recognition with ConvNet landmarks: Viewpoint-robust, condition-robust, training-free

Place recognition has long been an incompletely solved problem in that all approaches involve significant compromises. Current methods address many but never all of the critical challenges of place recognition – viewpoint-invariance, condition-invariance and minimizing training requirements. Here we present an approach that adapts state-of-the-art object proposal techniques to identify potential landmarks within an image for place recognition. We use the astonishing power of convolutional neural network features to identify matching landmark proposals between images to perform place recognition over extreme appearance and viewpoint variations. Our system does not require any form of training, all components are generic enough to be used off-the-shelf. We present a range of challenging experiments in varied viewpoint and environmental conditions. We demonstrate superior performance to current state-of-the- art techniques. Furthermore, by building on existing and widely used recognition frameworks, this approach provides a highly compatible place recognition system with the potential for easy integration of other techniques such as object detection and semantic scene interpretation.

“100 metros rasos”: objeto de aprendizagem para o ensino de física como aplicação do conectivismo e do edutretenimento

Pós-graduação em Televisão Digital: Informação e Conhecimento - FAAC

Generating Discriminative Objective Proposals via Submodular Ranking

Object recognition has long been a core problem in computer vision. To improve object spatial support and speed up object localization for object recognition, generating high-quality category-independent object proposals as the input for object recognition system has drawn attention recently. Given an image, we generate a limited number of high-quality and category-independent object proposals in advance and used as inputs for many computer vision tasks. We present an efficient dictionary-based model for image classification task. We further extend the work to a discriminative dictionary learning method for tensor sparse coding. In the first part, a multi-scale greedy-based object proposal generation approach is presented. Based on the multi-scale nature of objects in images, our approach is built on top of a hierarchical segmentation. We first identify the representative and diverse exemplar clusters within each scale. Object proposals are obtained by selecting a subset from the multi-scale segment pool via maximizing a submodular objective function, which consists of a weighted coverage term, a single-scale diversity term and a multi-scale reward term. The weighted coverage term forces the selected set of object proposals to be representative and compact; the single-scale diversity term encourages choosing segments from different exemplar clusters so that they will cover as many object patterns as possible; the multi-scale reward term encourages the selected proposals to be discriminative and selected from multiple layers generated by the hierarchical image segmentation. The experimental results on the Berkeley Segmentation Dataset and PASCAL VOC2012 segmentation dataset demonstrate the accuracy and efficiency of our object proposal model. Additionally, we validate our object proposals in simultaneous segmentation and detection and outperform the state-of-art performance. To classify the object in the image, we design a discriminative, structural low-rank framework for image classification. We use a supervised learning method to construct a discriminative and reconstructive dictionary. By introducing an ideal regularization term, we perform low-rank matrix recovery for contaminated training data from all categories simultaneously without losing structural information. A discriminative low-rank representation for images with respect to the constructed dictionary is obtained. With semantic structure information and strong identification capability, this representation is good for classification tasks even using a simple linear multi-classifier.

The European Union funded NEOShield project: A global approach to near-Earth object impact threat mitigation

Although discussions are underway within the Action Team 14 of the United Nations COPUOS, there is currently no concerted international plan addressing the impact threat from near-Earth objects (NEOs) and how to organize, prepare and implement mitigation measures. We report on a new international project to address impact hazard mitigation issues, being the subject of a proposal submitted to the European Commission in response to the 2011 FP7 Call "Prevention of impacts from near-Earth objects on our planet". Our consortium consists of 13 research institutes, universities, and industrial partners from 6 countries and includes leading US and Russian space organizations. The primary aim of the project, NEOShield, is to investigate in detail the three most promising mitigation techniques: the kinetic impactor, blast deflection,and the gravity tractor, and devise feasible demonstration missions. Furthermore, we will investigate options for an international strategy for implementation when an actual impact threat arises. The NEOShield project was formally accepted by the European Commission on 17 November 2011 and funded with a total of 5.8 million Euros for a period of 3.5 years. The kick-off meeting took place at the DLR Institute of Planetary Research, Berlin, in January 2012. In this paper we present a brief overview of the planned scope of the project.

View-Based Strategies for 3D Object Recognition

A persistent issue of debate in the area of 3D object recognition concerns the nature of the experientially acquired object models in the primate visual system. One prominent proposal in this regard has expounded the use of object centered models, such as representations of the objects' 3D structures in a coordinate frame independent of the viewing parameters [Marr and Nishihara, 1978]. In contrast to this is another proposal which suggests that the viewing parameters encountered during the learning phase might be inextricably linked to subsequent performance on a recognition task [Tarr and Pinker, 1989; Poggio and Edelman, 1990]. The 'object model', according to this idea, is simply a collection of the sample views encountered during training. Given that object centered recognition strategies have the attractive feature of leading to viewpoint independence, they have garnered much of the research effort in the field of computational vision. Furthermore, since human recognition performance seems remarkably robust in the face of imaging variations [Ellis et al., 1989], it has often been implicitly assumed that the visual system employs an object centered strategy. In the present study we examine this assumption more closely. Our experimental results with a class of novel 3D structures strongly suggest the use of a view-based strategy by the human visual system even when it has the opportunity of constructing and using object-centered models. In fact, for our chosen class of objects, the results seem to support a stronger claim: 3D object recognition is 2D view-based.

Using appearance and context for outdoor scene object classification

We propose a probabilistic object classifier for outdoor scene analysis as a first step in solving the problem of scene context generation. The method begins with a top-down control, which uses the previously learned models (appearance and absolute location) to obtain an initial pixel-level classification. This information provides us the core of objects, which is used to acquire a more accurate object model. Therefore, their growing by specific active regions allows us to obtain an accurate recognition of known regions. Next, a stage of general segmentation provides the segmentation of unknown regions by a bottom-strategy. Finally, the last stage tries to perform a region fusion of known and unknown segmented objects. The result is both a segmentation of the image and a recognition of each segment as a given object class or as an unknown segmented object. Furthermore, experimental results are shown and evaluated to prove the validity of our proposal

Object and Scene Classification: what does a Supervised Approach Provide us?

Given a set of images of scenes containing different object categories (e.g. grass, roads) our objective is to discover these objects in each image, and to use this object occurrences to perform a scene classification (e.g. beach scene, mountain scene). We achieve this by using a supervised learning algorithm able to learn with few images to facilitate the user task. We use a probabilistic model to recognise the objects and further we classify the scene based on their object occurrences. Experimental results are shown and evaluated to prove the validity of our proposal. Object recognition performance is compared to the approaches of He et al. (2004) and Marti et al. (2001) using their own datasets. Furthermore an unsupervised method is implemented in order to evaluate the advantages and disadvantages of our supervised classification approach versus an unsupervised one

The Dissociation of Location and Object Working Memory using fMRI and MEG

Visual working memory (VWM) involves maintaining and processing visual information, often for the purpose of making immediate decisions. Neuroimaging experiments of VWM provide evidence in support of a neural system mainly involving a fronto-parietal neuronal network, but the role of specific brain areas is less clear. A proposal that has recently generated considerable debate suggests that a dissociation of object and location VWM occurs within the prefrontal cortex, in dorsal and ventral regions, respectively. However, re-examination of the relevant literature presents a more robust distribution suggestive of a general caudal-rostral dissociation from occipital and parietal structures, caudally, to prefrontal regions, rostrally, corresponding to location and object memory, respectively. The purpose of the present study was to identify a dissociation of location and object VWM across two imaging methods (magnetoencephalography, MEG, and functional magnetic imaging, fMRI). These two techniques provide complimentary results due the high temporal resolution of MEG and the high spatial resolution of fMRI. The use of identical location and object change detection tasks was employed across techniques and reported for the first time. Moreover, this study is the first to use matched stimulus displays across location and object VWM conditions. The results from these two imaging methods provided convergent evidence of a location and object VWM dissociation favoring a general caudal-rostral rather than the more common prefrontal dorsal-ventral view. Moreover, neural activity across techniques was correlated with behavioral performance for the first time and provided convergent results. This novel approach of combining imaging tools to study memory resulted in robust evidence suggesting a novel interpretation of location and object memory. Accordingly, this study presents a novel context within which to explore the neural substrates of WM across imaging techniques and populations.

Resolution limits to object tracking with subpixel accuracy

Subpixel methods increase the accuracy and efficiency of image detectors, processing units, and algorithms and provide very cost-effective systems for object tracking. Published methods achieve resolution increases up to three orders of magnitude. In this Letter, we demonstrate that this limit can be theoretically improved by several orders of magnitude, permitting micropixel and submicropixel accuracies. The necessary condition for movement detection is that one single pixel changes its status. We show that an appropriate target design increases the probability of a pixel change for arbitrarily small shifts, thus increasing the detection accuracy of a tracking system. The proposal does not impose severe restriction on the target nor on the sensor, thus allowing easy experimental implementation.

Three-dimensional planar object tracking with sub-pixel accuracy

Subpixel techniques are commonly used to increase the spatial resolution in tracking tasks. Object tracking with targets of known shape permits obtaining information about object position and orientation in the three-dimensional space. A proper selection of the target shape allows us to determine its position inside a plane and its angular and azimuthal orientation under certain limits. Our proposal is demonstrated both numerical and experimentally and provides an increase the accuracy of more than one order of magnitude compared to the nominal resolution of the sensor. The experiment has been performed with a high-speed camera, which simultaneously provides high spatial and temporal resolution, so it may be interesting for some applications where this kind of targets can be attached, such as vibration monitoring and structural analysis.

3D Visual Data-Driven Spatiotemporal Deformations for Non-Rigid Object Grasping Using Robot Hands

Sensing techniques are important for solving problems of uncertainty inherent to intelligent grasping tasks. The main goal here is to present a visual sensing system based on range imaging technology for robot manipulation of non-rigid objects. Our proposal provides a suitable visual perception system of complex grasping tasks to support a robot controller when other sensor systems, such as tactile and force, are not able to obtain useful data relevant to the grasping manipulation task. In particular, a new visual approach based on RGBD data was implemented to help a robot controller carry out intelligent manipulation tasks with flexible objects. The proposed method supervises the interaction between the grasped object and the robot hand in order to avoid poor contact between the fingertips and an object when there is neither force nor pressure data. This new approach is also used to measure changes to the shape of an object’s surfaces and so allows us to find deformations caused by inappropriate pressure being applied by the hand’s fingers. Test was carried out for grasping tasks involving several flexible household objects with a multi-fingered robot hand working in real time. Our approach generates pulses from the deformation detection method and sends an event message to the robot controller when surface deformation is detected. In comparison with other methods, the obtained results reveal that our visual pipeline does not use deformations models of objects and materials, as well as the approach works well both planar and 3D household objects in real time. In addition, our method does not depend on the pose of the robot hand because the location of the reference system is computed from a recognition process of a pattern located place at the robot forearm. The presented experiments demonstrate that the proposed method accomplishes a good monitoring of grasping task with several objects and different grasping configurations in indoor environments.

Towards a Research Proposal Outline of the Central Library

Few studies relating to librarianship as a science, because it is argued that the Library lacks a body of theory, an object of study and methodology of its own. According to Murcia and Tamayo "science is a critical task not dogmatic, that puts all its cases to trial and criticism." 0 is, science tests the presumed knowledge. He adds that science related facts and interconnected with each other, in order to achieve logical connections that allow the provision of postulates and axioms from the systematization achieved through research and the scientific method to determine the objectivity that exists between facts and phenomena. For him, scientific research is the instrument by which science makes it to scientific knowledge. (1982, p. 11)

Experiences with an Object-Oriented, Multi-Stage Language

Metaphor is a multi-stage programming language extension to an imperative, object-oriented language in the style of C# or Java. This paper discusses some issues we faced when applying multi-stage language design concepts to an imperative base language and run-time environment. The issues range from dealing with pervasive references and open code to garbage collection and implementing cross-stage persistence.