Dynamic perception after visually guided grasping by a human-like autonomous robot

We will explore dynamic perception following the visually guided grasping of several objects by a human-like autonomous robot. This competency serves for object categorization. Physical interaction with the hand-held object gives the neural network of the robot the rich, coherent and multi-modal sensory input. Multi-layered self-organizing maps are designed and examined in static and dynamic conditions. The results of the tests in the former condition show its capability of robust categorization against noise. The network also shows better performance than a single-layered map does. In the latter condition we focus on shaking behavior by moving only the forearm of the robot. In some combinations of grasping style and shaking radius the network is capable of categorizing two objects robustly. The results show that the network capability to achieve the task largely depends on how to grasp and how to move the objects. These results together with a preliminary simulation are promising toward the self-organization of a high degree of autonomous dynamic object categorization.     

Visually guided grasping in unstructured environments

We present simple and robust algorithms which combine uncalibrated stereo vision and a robot manipulator to enable it locate, reach and grasp unmodelled objects in unstructured environments. In the first stage, an operator indicates the object to be grasped by simply pointing at it. Next, the vision system segments the indicated object from the background, and plans a suitable grasp strategy. Finally, the robotic arm reaches out towards the object and executes the grasp. Uncalibrated stereo vision allows the system to continue to operate in the presence of errors in the kinematics of the robot manipulator and unknown changes in the position, orientation and intrinsic parameters of the stereo cameras during operation.     

Evolution of visually guided behavior in artificial agents

Recent work on brightness, color, and form has suggested that human visual percepts represent the probable sources of retinal images rather than stimulus features as such. Here we investigate the plausibility of this empirical concept of vision by allowing autonomous agents to evolve in virtual environments based solely on the relative success of their behavior. The responses of evolved agents to visual stimuli indicate that fitness improves as the neural network control systems gradually incorporate the statistical relationship between projected images and behavior appropriate to the sources of the inherently ambiguous images. These results: (1) demonstrate the merits of a wholly empirical strategy of animal vision as a means of contending with the inverse optics problem; (2) argue that the information incorporated into biological visual processing circuitry is the relationship between images and their probable sources; and (3) suggest why human percepts do not map neatly onto physical reality.     

Discrete event modeling of visually guided behaviors

When visual behaviors are combined to provide a specific functionality needed for a task, the combination is often based on heuristic rules. In this article we show that by adopting the Discrete-Event Systems (DES) formalism for describing the interaction between visual behaviors it is possible to provide systems that have well-defined properties in terms of observability and controllability. The method is in particular suited for describing the coupling between action and perception. An introduction to the use of DES is provided and it is demonstrated how DES are used for modeling behaviors and controlling a mobile robot equipped with a binocular camera head and some additional sensors.     

Right hand advantage in visually guided reaching and aiming movements: brief review and comments

Although understanding of the organization and control of visually guided reaching and aiming movements is still sketchy and incomplete, evidence from behavioural studies supports the contention that right-handed individuals typically execute aiming movements with better speed, smoothness and consistency, and with a greater degree of spatial precision when performing them with their right hand. Creative attempts to account for the superiority of the right hand on a variety of visually guided reaching and aiming tasks have focused on the processing characteristics of the contralateral or left cerebral hemisphere. This brief review summarizes the research conducted over the last few decades on the subject, highlights the theoretical interpretations offered to explain manual asymmetries in the organization and control of goal-directed movements and identifies directions for further empirical research. The theoretical and practical implications of laterality research efforts along the lines of goal-directed behaviour are discussed.     

Right hand advantage in visually guided reaching and aiming movements: brief review and comments

Although understanding of the organization and control of visually guided reaching and aiming movements is still sketchy and incomplete, evidence from behavioural studies supports the contention that right-handed individuals typically execute aiming movements with better speed, smoothness and consistency, and with a greater degree of spatial precision when performing them with their right hand. Creative attempts to account for the superiority of the right hand on a variety of visually guided reaching and aiming tasks have focused on the processing characteristics of the contralateral or left cerebral hemisphere. This brief review summarizes the research conducted over the last few decades on the subject, highlights the theoretical interpretations offered to explain manual asymmetries in the organization and control of goal-directed movements and identifies directions for further empirical research. The theoretical and practical implications of laterality research efforts along the lines of goal-directed behaviour are discussed.     

Effects of increasing visual load on aurally and visually guided target acquisition in a virtual environment

The aim of the present study is to investigate interactions between vision and audition during a target acquisition task performed in a virtual environment. We measured the time taken to locate a visual target (acquisition time) signalled by auditory and/or visual cues in conditions of variable visual load. Visual load was increased by introducing a secondary visual task. The auditory cue was constructed using virtual three-dimensional (3D) sound techniques. The visual cue was constructed in the form of a 3D updating arrow. The results suggested that both auditory and visual cues reduced acquisition time as compared to an uncued condition. Whereas the visual cue elicited faster acquisition time than the auditory cue, the combination of the two cues produced the fastest acquisition time. The introduction of secondary visual task differentially affected acquisition time depending on cue modality. In conditions of high visual load, acquiring a target signalled by the auditory cue led to slower and more error-prone performance than acquiring a target signalled by either the visual cue alone or by both the visual and auditory cues.     

Learning without lectures: a case study

During a 1995-1996 sabbatical, the authors' long-standing concern about discrepancies between their teaching philosophies and classroom practices reached a crisis. As a result, they decided to offer the fall 1996 Introduction to Psychology course on the World Wide Web. There were several issues to think about. First, the authors saw the potential for a huge impact, since 80 to 85 percent of North Dakota State University students enrol in Psych110, most within their first two years. Second, they decided to design the course so that it could be offered for off-campus distance learning as well as on-campus resident instruction. Third, they decided to minimize overhead for students by using only a Web browser, e-mail, a textbook, and the textbook publisher's demonstration/simulation software. Fourth, they chose to base decisions about course components on published literature covering cognition and teaching. Finally, they built in mid term and final student course evaluation components and a field experiment comparing student attitudes and learning outcome for online versus lecture classes     

A Bayesian exploration-exploitation approach for optimal online sensing and planning with a visually guided mobile robot

We address the problem of online path planning for optimal sensing with a mobile robot. The objective of the robot is to learn the most about its pose and the environment given time constraints. We use a POMDP with a utility function that depends on the belief state to model the finite horizon planning problem. We replan as the robot progresses throughout the environment. The POMDP is high-dimensional, continuous, non-differentiable, nonlinear, non-Gaussian and must be solved in real-time. Most existing techniques for stochastic planning and reinforcement learning are therefore inapplicable. To solve this extremely complex problem, we propose a Bayesian optimization method that dynamically trades off exploration (minimizing uncertainty in unknown parts of the policy space) and exploitation (capitalizing on the current best solution). We demonstrate our approach with a visually-guide mobile robot. The solution proposed here is also applicable to other closely-related domains, including active vision, sequential experimental design, dynamic sensing and calibration with mobile sensors.     

Learning object,grasping and manipulation activities using hierarchical HMMs

This article presents a probabilistic algorithm for representing and learning complex manipulation activities performed by humans in everyday life. The work builds on the multi-level Hierarchical Hidden Markov Model (HHMM) framework which allows decomposition of longer-term complex manipulation activities into layers of abstraction whereby the building blocks can be represented by simpler action modules called action primitives. This way, human task knowledge can be synthesised in a compact, effective representation suitable, for instance, to be subsequently transferred to a robot for imitation. The main contribution is the use of a robust framework capable of dealing with the uncertainty or incomplete data inherent to these activities, and the ability to represent behaviours at multiple levels of abstraction for enhanced task generalisation. Activity data from 3D video sequencing of human manipulation of different objects handled in everyday life is used for evaluation. A comparison with a mixed generative-discriminative hybrid model HHMM/SVM (support vector machine) is also presented to add rigour in highlighting the benefit of the proposed approach against comparable state of the art techniques.     

Automatic test case optimization: a bacteriologic algorithm

Improving test cases automatically is a nonlinear optimization problem. To solve this problem, we've developed a bacteriologic algorithm, adapted from genetic algorithms that can generate and optimize a set of test cases. A .NET component that parses C# source files illustrates our algorithm.     

Semantics guided regression test cost reduction

Software maintainers are faced with the task of regression testing: retesting a modified program on an often large number of test cases. The cost of regression testing can be reduced if the size of the program is reduced and if old test cases and results can be reused. Two complimentary algorithms for reducing the cost of regression testing are presented. The first produces a program called Differences that captures the semantic change between Certified, a previously tested program, and Modified, a changed version of Certified. It is more efficient to test Differences, because it omits unchanged computations. The program Differences is computed using a combination of program slices. The second algorithm identifies test cases for which Certified and Modified produce the same output and existing test cases that test new components in Modified. The algorithm is based on the notion of common execution patterns. Program components with common execution patterns have the same execution pattern during some call to their procedure. They are computed using a calling context slice. Whereas an interprocedural slice includes the program components necessary to capture all possible executions of a statement, a calling context slice includes only those program components necessary to capture the execution of a statement in a particular calling context. Together with Differences, it is possible to test Modified by running Differences on a smaller number of test cases. This is more efficient than running Modified on a large number of test cases. A prototype implementation has been built to examine and illustrate these algorithms     

Automatic test generation: a use case driven approach

Use cases are believed to be a good basis for system testing. Yet, to automate the test generation process, there is a large gap to bridge between high-level use cases and concrete test cases. We propose a new approach for automating the generation of system test scenarios in the context of object-oriented embedded software, taking into account traceability problems between high-level views and concrete test case execution. Starting from a formalization of the requirements based on use cases extended with contracts, we automatically build a transition system from which we synthesize test cases. Our objective is to cover the system in terms of statement coverage with those generated tests: an empirical evaluation of our approach is given based on this objective and several case studies. We briefly discuss the experimental deployment of our approach in the field at Thales Airborne Systems.     

Learning discriminative features with a dual-constrained guided network for video-based person re-identification

Multimedia Tools and Applications - Video-based person re-identification (ReID) aims at matching pedestrians in a large video gallery across different cameras. However, some interference factors in...     

Learning in multi-agent systems: a case study of construction claims negotiation

The ability of agents to learn is of growing importance in multi-agent systems. It is considered essential to improve the quality of peer to peer negotiation in these systems. This paper reviews various aspects of agent learning, and presents the particular learning approach—Bayesian learning—adopted in the MASCOT system (multi-agent system for construction claims negotiation). The core objective of the MASCOT system is to facilitate construction claims negotiation among different project participants. Agent learning is an integral part of the negotiation mechanism. The paper demonstrates that the ability to learn greatly enhances agents' negotiation power, and speeds up the rate of convergence between agents. In this case, learning is essential for the success of peer to peer agent negotiation systems.     

Artificial intelligence test: a case study of intelligent vehicles

To meet the urgent requirement of reliable artificial intelligence applications, we discuss the tight link between artificial intelligence and intelligence test in this paper. We highlight the role of tasks in intelligence test for all kinds of artificial intelligence. We explain the necessity and difficulty of describing tasks for intelligence test, checking all the tasks that may encounter in intelligence test, designing simulation-based test, and setting appropriate test performance evaluation indices. As an example, we present how to design reliable intelligence test for intelligent vehicles. Finally, we discuss the future research directions of intelligence test.     

The revealing power of a test case

‘Propagation, infection, and execution analysis’ (termed PIE) is used for predicting where faults can more easily hide in software. To make such predictions, programs are dynamically executed with test cases, and information concerning the test cases is collected into a histogram, each bin of which represents a single test case. The score in a bin predicts the likelihood that the test case will reveal a fault through the production of a failure (if a fault exists in the set of program locations that the test case executes). Preliminary experiments using program mutations suggest that the histogram technique presented in this paper can rank test cases according to their fault revealing ability.     

Model-based test case generation and prioritization: a systematic literature review

Software and Systems Modeling - Model-based test case generation (MB-TCG) and prioritization (MB-TCP) utilize models that represent the system under test (SUT) for test generation and...