Search

Humans use information from sensory predictions, together withcurrent observations, for the optimal exploration and recognition oftheir surrounding environment. In this work, two novel adaptiveperception strategies are proposed for accurate and fast exploration ofobject shape with a robotic tactile sensor. These strategies called 1)adaptive weighted prior and 2) adaptive weighted posterior, combinetactile sensory predictions and current sensor observations toautonomously adapt the accuracy and speed of active Bayesian perceptionin object exploration tasks. Sensory predictions, obtained from a forwardmodel, use a novel Predicted Information Gain method. These predictionsare used by the tactile robot to analyse `what would have happened' ifcertain decisions `would have been made' at previous decision times. Theaccuracy of predictions is evaluated and controlled by a confidenceparameter, to ensure that the adaptive perception strategies rely more onpredictions when they are accurate, and more on current sensoryobservations otherwise. This work is systematically validated with therecognition of angle and position data extracted from the exploration ofobject shape, using a biomimetic tactile sensor and a robotic platform.The exploration task implements the contour following procedure used byhumans to extract object shape with the sense of touch. The validationprocess is performed with the adaptive weighted strategies and activeperception alone. The adaptive approach achieved higher angle accuracy(2.8 deg) over active perception (5 deg). The position accuracy wassimilar for all perception methods (0.18 mm). The reaction time or numberof tactile contacts, needed by the tactile robot to make a decision, wasimproved by the adaptive perception (1 tap) over active perception (5taps). The results show that the adaptive perception strategies canenable future robots to adapt their performance, while improving thetrade-off between accuracy and reaction time, for tactile exploration,interaction and recognition tasks.

In the construction industry, rebar crosspoints binding relies heavily on manual work, which has become the bottleneck of construction-efficiency improvement. In this study, a full-automatic robot system, based on active perception and planning, is developed to realize the automation of the rebar crosspoints binding process. Based on the preprocessed image from an RGBD camera, a neural network method is proposed to recognize the rebar crosspoints. An active planning method to traverse the rebar plane is designed in the results of crosspoints recognition. Experiment results show that the rebar crosspoints recognition method has high accuracy (the detection accuracy is more than 89% and the classification accuracy is more than 98%). Experiments in realistic scenarios show that the robot system can traverse the rebar plane and bind the rebar crosspoints automatically to reduce labor costs. In the future, the robot system will work in curved environments and have higher detection accuracy.

Many robots are limited in their operating capabilities, both computational and energy-wise. A strong desire exists to keep computation cost and energy consumption to a minimum when executing tasks like object recognition with a mobile robot. Adaptive action selection is a paradigm, offering great flexibility in trading off the cost of acquiring information against making robust and reliable inference under uncertainty. In this paper, we study active multi-view object recognition and describe an information-theoretic framework that combines and unifies two common techniques: online feature selection for reducing computational costs and view planning for resolving ambiguities and occlusions. Our algorithm adaptively chooses between the two strategies of either selecting only the features that are most informative to the recognition, or moving to a new viewpoint that optimally reduces the expected uncertainty on the identity of the object. This two step process allows us to keep overall computation cost minimal but simultaneously increase recognition accuracy. Extensive empirical studies on a large RGB-D dataset, and with two different feature sets, have validated the effectiveness of the proposed framework. Our experiments show that dynamic feature selection alone reduces the computation time at runtime 2.5-6 times and, when combining it with viewpoint selection, we significantly increase the recognition accuracy on average by 8%-18% absolute compared to systems that do not use these two strategies. By establishing a link between active object recognition and change detection, we were further able to use our framework for the follow-up task of actively detecting object change. Furthermore, we have successfully demonstrated the framework's applicability to a low-powered quadcopter platform with limited operating time. A unified approach to feature and viewpoint selection for multi-view object recognition is proposed.Online feature selection reduces the dimensionality and with that the computation time.View planning offers performance advantages whenever multiple views are required due to ambiguous situations or occlusions.Increased recognition accuracy and reduced computation cost are realized by an information-theoretic action selection framework.

This article proposes a new theoretical framework of environmental satisfaction and human comfort, highlighting the role of human agency in the interaction between humans and environment and the context within which such interactions occur. The framework is constructed by building on Bronfenbrenner's bioecological model and logically connecting existing theories in the broader area of social sciences. The proposed framework conceptualizes satisfaction and comfort as an optimization process accomplished through one's active perception, interpretation, and modification of his/her socio-physical environment. This optimization process is viewed as ongoing and cyclical. Four modes of optimization are proposed: environmental modification, behavioral adaptation, normative adaptation, and withdrawal. Three methodological implications are then discussed: using multiple sources of information to examine divergent worldviews amongst involved social groups; building case study profiles for each place type using the proposed theoretical framework; and employing multi-level, interdisciplinary approaches in both research and practices.

In this paper, we present a system for autonomous object search and exploration in cluttered environments. The system shortens the average time needed to complete search tasks by continually planning multiple perception actions ahead of time using probabilistic prior knowledge. Useful sensing actions are found using a frontier-based view sampling technique in a continuously built 3D map. We demonstrate the system on real hardware, investigate the planner’s performance in three experiments in simulation, and show that our approach achieves shorter overall run times of search tasks compared to a greedy strategy.

Discriminating or classifying different terrains is an important ability for every autonomous mobile robot. A variety of sensors, preprocessing techniques, and algorithms in different robots were applied. However, little attention was paid to the way sensory data was generated and to the contribution of different sensory modalities. In this work, a quadruped robot traversing different grounds using a variety of gaits is used, equipped with a collection of proprioceptive (encoders on active, and passive compliant joints), inertial, and foot pressure sensors. The effect of different gaits on classification performance is assessed and it is demonstrated that separate terrain classifiers for each motor program should be employed. Furthermore, poor performance of randomly generated motor commands confirms the importance of coordinated behavior on sensory information structuring. The collection of sensors sensitive to active, "tactile", terrain exploration proved effective. Among the individual modalities, encoders on passive compliant joints delivered best results. Preconditioning on gait used boosts terrain discrimination in a legged robot.Specific gaits are particularly suited for terrain perception.Inertial, tactile, and proprioceptive sensors are a robust terrain sensing set.Encoders in passive compliant joints performed best from the sensory set.

Consideration is given to the hypothesis provided by Trevarthen and Delafield-Butt (2013) – that disruption of social synchrony in autism is rooted in disturbance of prospective motor control for expressive action and active perception of other's expression: crucially, autism is a disturbance of “sensorimotor intentionality” (Delafield-Butt and Gangopadhyay 2013). The feedforward qualities of communicative musicality ( Malloch, Trevarthen, Malloch, and Trevarthen (2009)) ) – the temporal motivation and organisation inherent in pulse and vitality – are taken as foundational for a series of therapeutic principles developed here. The aim: to support disrupted prosocial “intentional movements” by overlaying, integrating, and piggy-backing on intact rhythm and vitality imported, in the moment, from cross-modality experiences – to imbue a child’s interaction with forward looking organisation and impulse to interact. These principles are illustrated in the form of mini case studies, taken from real play therapy sessions. Certain key concepts are introduced, including: the use of rhythm as compelling structure, the 'jazz gap' as temporal catalyst, the calming and engaging effect of inter-synchrony, and the vitality-matching of a child's emotive and motoric patterns.

In this article we present an approach to the design of human-like artificial systems. It uses a perception model to describe how sensory information is processed for a particular task and to correlate human and artificial perception. Since human-like sensors share their principle of operation with natural systems, their response can be interpreted in an intuitive way. Therefore, such sensors allow for easier and more natural human-machine interaction. The approach is demonstrated in two applications. The first is an ''electronic tongue'', which performs quality assessment of food and water. In the second application we describe the development of an artificial hand for dexterous manipulation. We show that human-like functionality can be achieved even if the structure of the system is not completely biologically inspired.

Olfaction is a challenging new sensing modality for intelligent systems. With the emergence of electronic noses, it is now possible to detect and recognize a range of different odours for a variety of applications. In this work, we introduce a new application where electronic olfaction is used in cooperation with other types of sensors on a mobile robot in order to acquire the odour property of objects. We examine the problem of deciding when, how and where the electronic nose (e-nose) should be activated by planning for active perception and we consider the problem of integrating the information provided by the e-nose with both prior information and information from other sensors (e.g., vision). Experiments performed on a mobile robot equipped with an e-nose are presented. © 2005 Elsevier B.V. All rights reserved.

This paper outlines a philosophical and psycho-physiological basis for embodied perception, and develops a framework for conceptual embodiment of vision-guided robots. We argue that categorisation is important in all stages of robot vision. Further, that classical computer vision is unsuitable for this categorisation, however, through conceptual embodiment, active perception can be effective. We present a methodology for developing vision-guided robots that applies embodiment, explicitly and implicitly, in categorising visual data to facilitate efficient perception and action. Finally, we present systems developed using this methodology, and demonstrate that embodied categorisation can make algorithms more efficient and robust.

In this paper, we propose an image segmentation system adapted to the uniform and/or weakly textured region extraction. The architecture of the proposed system combines two concepts. (i) The integration of the information resulting from two complementary segmentation methods: edge detection and region extraction. Thus, this allows us to exploit the advantages of each. (ii) The active perception via the intermediate of a feedback. This permits the correction and adjustment of the control parameters of the methods used. The originality of the suggested co-operation carries on the introduction of a mechanism, which checks the coherence of the results through a comparison of the two segmentations. From over-segmentation results, both methods are iterated by loosening certain constraints, until they converge towards stable and coherent results. This coherence is achieved by minimising a dissimilarity measure between the edges and the boundaries of the regions. The aim is therefore to provide the optimal solution in the sense of compatibility between the segmentation results. The system therefore uses a hybrid co-operation approach and is almost automatic and unsupervised. The performance of this approach has been measured on two remote sensing applications: agricultural landscape segmentation and forestry vegetation classification. q 2002 Published by Elsevier Science B.V.

The potential of airborne sonar to provide effective information about three-dimensional (3D) spatial layouts was assessed in four companion experiments. Blindfolded participants, never given visual access to the layout of a large room, were asked to use a sonar device whose output they had never previously encountered to judge the passability (by normal walking) of apertures between two aligned wall panels. Estimates were made from fixed and variable locations, approaches to the apertures were made from orthogonal and oblique angles, and the panels were at different distances and orientations. In each experiment, participants gave evidence of an ability to immediately use the information in structured echoes to make these judgments, though aperture location, approach angles, wall alignment and orientation each had significant effects on performance. The data are compared with performance under visual and nonechoic auditory conditions and are discussed with respect to the notions of potential information and effective information during these perceptually guided tasks.

The purpose of this paper is to introduce the reader to a novel approach to data fusion. We focus on the latest results which have immediate practical implications. Many tasks in active perception require the ability to combine information from a variety of sensors. Prior to combination, the data must be tested for consistency. Both of these tasks can be viewed as data fusion problems. We examine such problems for location data models. Our approach is based on statistical decision theory. We present the application of the theory to mobile robot localization.

This paper presents a representation for objects based not only on shape and material properties but also on the role that the object plays in the scene. Functionality is introduced as a representation mapping the intrinsic properties of an object in the perceptual and interactive space of an agent and is characterized as an extension of purposive and active vision, active perception and task-oriented approaches. The developed robotic system extracts functional features employed in interactions for piercing and chopping. The representation is presented by means of histograms and Force–Shape maps. Histograms are provided as means of representing the success in accomplishing the various experiments utilizing different tools and materials. Force–Shape maps are employed to express the dynamics of functional interactions. The approach described in this paper has two goals: a description of the components constituting a representation of functionality and a methodology for a robotic agent to extract a representation.

This article describes a method for learning a set of landmarks suitable for place navigation. The approach is novel in that it exploits the ability of a robot to learn through active perception in the task environment, similar to the learning by experimentation technique developed for LEX (Mitchell et al., 1990). The proposed strategy uses heuristics to select and rank candidate triples, then generates test cases to confirm that the best triple is sufficient. The method supports the use of multiple sensors with different computational and energy costs, where a utility function captures the tradeoff between navigational performance ranking and cost. Over 100 data points were collected on a mobile robot using a laser barcode reader and computer vision to identify landmarks. The results indicated that active perception and experimentation identified triples with better navigational properties. Furthermore, the learning process is proactive: it was shown to prevent the robot from learning a triple which was not visible over the entire navigational space and/or was not sufficient in practice.

This paper describes a novel approach in target recognition, using active and selective perception and perceptual learning in the analysis of time-varying imagery. Attention mechanisms are implemented through the use of three linked functional pyramids corresponding to image representation, memory, and attention. Multiresolution search, supported by saccade and zoom-lens control, allows a system's focus of attention to roam inside an image pyramid under the guidance of an attention pyramid and analyse a field-of-view at any available resolution. Simulations showed this approach reducing the volume of data to be processed and performing robustly as characterized by the absence of false negatives and the presence of very few false positives.

Functionality of an object defines its applicability to a task, In this paper, we introduce a representation for functionality and present a methodology for its recovery. In addition, we introduce force-shape maps as means to integrate and classify the recovered functionality, Since functionality describes an interaction, the representation for an object must include not only its intrinsic (material) but also its functional (how it is used) properties. In order to recover the functionality of an object, we have developed a formalism based on discrete event system theory and on the paradigm of active perception. This formalism allows us to express an investigating task in the form of a finite automaton for controlling and observing the interactions in our task. Our current investigation focuses on manipulatory interactions, and thus far, the functionality of piercing. We envision an expandable system having at its disposal a score of investigative procedures and through them to be able to classify and recognize an object based on its functional attributes.

This paper presents a methodology for investigating the functionality of objects in manipulatory tasks. Functionality is defined as the applicability of an object for the fulfillment of a particular purpose. The description of an interaction for investigating the functionality of an object is described using a formalism based on Discrete Event System Theory and the Active Perception paradigm. The notion of an overseer for controlling, arbitrating and fusing assertions from multiple observers is presented. To describe the overall task observability we introduce the concept of piecewise observability. The functionality of a tool for piercing is presented and the feasibility of approach is demonstrated in a series of experiments using a roboti system and a compliant force sensor.

The main goal of this paper is to show that there is a natural flow from active perception through exploration to perceptual learning. W have attempted to conceptualize the perceptual process of an organism that has the top-level task of surviving in an unknown environment. During this conceptualization process. Four necessary ingredients have emerged for either artificial or biological organisms. First, the sensory apparatus and processing of the organism must be active and flexible. Second, the organism must have exploratory capabilities. Third, the organism must be selective in its data acquisition process. Fourth, the organism must be able to learn. In the section on learning, we have clearly delineated the difference between what must be innate and what must be learned. In order to test our theory, we present the system's architecture that follows from the perceptual task decomposition. The predictions of this theory are that an artificial system can explore and learn about its environment modulo its sensors, manipulators, end effectors, and exploratory procedures/attribute extractors. It can describe its world with respect to the built-in alphabet, that is, the set of perceptual primitives.

Tan M. and Schlimer, J.C., A cost-sensitive machine learning method for the approach and recognize task, Robotics and Autonomous Systems, 8 (1991) 31–45. The long-term goal of this reasearch is to build a robot which can survive in an unanticipated environment for a long time. Such a robot must learn which sensors to use, where to use them, and how to generate inexpensive and reliable effector procedures to accomplish its tasks. Current machine learning methods are inadequate because they ignore execution costs and are ill-prepared to handle prediction failures. This paper described the CSI, learning and control system that attempts to address this goal, CSL has been implemented in a real-world robot for the approach and recognize taks.

If all of the motive forces of the physical world, physically of the world, can be agreed upon to occur from the conceptual/ethereal as a confrontation of the logically entailed with its' antithesis, as the paradox, a scheme might be evolved to detail all phenomenon as occurrences that are the result of an elemental universal construction effected by the inversion, then one might account for the sciences, its' theory, the questions of nature, natural history, human history, rather than as explanation, as if seeking explanation for existence in the third person, but as an accounting that is closer conceptually to the innately perceived. If it is at the juncture of paradox where the pursuits of civilization seem to expand in a never ending seeking, perhaps one might reorient his frame to define in terms of paradox. Though this reorientation does not appear on the surface as potentially productive, does not resolve the paradox, one might assume it to be the only route available and, if not productive, one might also potentially assume that all efforts to extract a knowledge of nature are in vain in the sense that one might only arrive at paradox, originating itself from an active pursuit to bear fruit for continuance. At the surface of this inquiry is presentation of a valid criteria for the ubiquitous existence of paradox throughout all of nature, though more in line with presentation first of paradox and subsequently exposed as ubiquitous and employed for explanation. Since paradox arises from the willed/active application of the intellect to the external world there is no ground to assume that it is innate itself in a simpler setting of nature, the natural world and its' processes, in this presentation I will try to establish that paradox, defined as logically irreconcilable conflict, other than being solely the product of active perception, is, in an analogous corresponding form, ubiquitously companioned ,and tangential to direction, a partner to the inversion that pervades the universe. The inversion is given as a substrate for the faces of the paradox and the two combined as an elemental unit of nature. With regards to paradoxes of mind and matter, an ultimate conceptually/philosophically arrived paradox, a synthesis of the physical/biological/philosophical concepts with the empirical, DNA is given a role as the actual embodiment of all that can and does ensue in the universe - path and memory. In light of theory and current research results in the science of physics, DNA is elevated, with potential, scientific validity, to possess both a ubiquitous nature and unique existence confined to biological entities as 'a physical piece of physical path', a biological source of all memory, perception and cognition, a physically energetic absolute and fixed differential standard of path/emergence in physical form, contained physical emergence that serves to energetically emerge, and defining of all from a first perspective beholding to the possessor of life as a sole criteria for perspective and reference with which to scientifically organize experiences of nature.

Rodents demonstrate an outstanding capability of tactile perception with their whiskers. Mechanoreceptors surrounding the whisker shaft in their follicle structure measure deflection of the whisker. We designed biomimetic whiskers following the basic design of the follicle. In experiments with the artificial whiskers, we have explored tactile perception based on active whisking where the deflection angle or velocity provides the localization information which is the basis of shape recognition. Measuring contact distances at varying protraction angles allows discrimination of round objects with a varying curvature, or objects with different lateral shapes, such as square and round objects. We show the capabilities and limitations of a single whisker for shape recognition as well as the usefulness of multiple whiskers. In addition, measuring both vertical and horizontal deflection of a single whisker allows detection of the vertical shape for objects with a smooth surface. Two or more whiskers stacked vertically can recognize the vertical shape by observing the difference of their deflection amplitudes or the time shift of deflection velocity peak. The results provide a clue on how autonomous robots could improve their sensory capabilities with mechanical probes.

The paper discusses two different kinds of disturbances of the active visual perception of complex patterns: using the recording of ocular movements during 2–3 min of observation of complex pictures. In normal persons a preliminary period of orientation in complex visual patterns can be seen, and a final conclusion as to the meaning of the picture is a result of complicated information processing. In patients with lesions of parieto-occipital zones of the brain attempts at preliminary active orientation are preserved, but the evaluation of the visual patterns suffers in its ultimate executive phase. In patients with severe bilateral lesions of the frontal lobes the preliminary active orientation in visual patterns suffers, eye movements are of a stereotype character and evaluation of the complex visual picture does not follow a complex orientation-investigation period.