Success rates in a multimodal command language for home robot users

This article considers the success rates in a multimodal command language for home robot users. In the command language, the user specifies action types and action parameter values to direct robots in multiple modes such as speech, touch, and gesture. The success rates of commands in the language can be estimated by user evaluations in several ways. This article presents some user evaluation methods, as well as results from recent studies on command success rates. The results show that the language enables users without much training to command home robots at success rates as high as 88%–100%. It is also shown that multimodal commands combining speech and button-press actions included fewer words and were significantly more successful than single-modal spoken commands.     

Characterizing user performance in command-driven dialogue

To learn to use an interactive system, a person typically has to acquire a good deal of new knowledge. The ease of learning will depend on the extent to which the design of the task and the interface capitalizes on the user's pre-existing knowledge and his or her cognitive capabilities for learning. This paper explores the nature of both design decisions and user learning with a command-based system. Three studies were conducted, all involving a task in which secret messages were decoded by means of a sequence of commands (based on the task used by Barnard et al. In Study I, software specialists designed command structures for the task and gave reasons for their choices. In Study II, naive subjects chose between alternative command terms. In Study TTI, subjects learned to use interactive versions of the task in which dialogue factors (command terms and argument structures) were systematically varied. The results enabled the development of user knowledge of the system to be specified in detail. Comparisons across the three studies highlighted the diversity of the factors determining both design decisions and user behaviour.     

Understanding user commands by evaluating fuzzy linguistic information based on visual attention

This article proposes a method for understanding user commands based on visual attention. Normally, fuzzy linguistic terms such as “very little” are commonly included in voice commands. Therefore, a robot’s capacity to understand such information is vital for effective human-robot interaction. However, the quantitative meaning of such information strongly depends on the spatial arrangement of the surrounding environment. Therefore, a visual attention system (VAS) is introduced to evaluate fuzzy linguistic information based on the environmental conditions. It is assumed that the corresponding distance value for a particular fuzzy linguistic command depends on the spatial arrangement of the surrounding objects. Therefore, a fuzzy-logic-based voice command evaluation system (VCES) is proposed to assess the uncertain information in user commands based on the average distance to the surrounding objects. A situation of object manipulation to rearrange the user’s working space is simulated to illustrate the system. This is demonstrated with a PA-10 robot manipulator.     

A theory of command language dialogue for a knowledge‐based human‐computer interaction

The flexibility and usability of graphic‐based HCIs can be increased by adding a natural language interface with command menus. Among the several other advantages, such embellishment offers the user an opportunity for direct expression of his or her behaviors, goals, intentions, and objectives along the continuum of the task knowledge. The existing graphic‐based HCIs that operate on active symbologies and icons assume the user's mental models to correlate with perceptual and cognitive levels of the task understanding. This obviously increases mental loads and the frustration of the human adapting to the system. In reality, the system should be designed to adapt to the user's behavior and skill level. In order to improve the current design of graphic‐based HCIs, we have formulated theories of command production language that will enhance the user's ability to interact with the system. The methods developed combine the theory of expert database with formal grammar to develop command‐production rules using a natural language dictionary prototype. We show that the commands are linear, regular, and symmetric, although conforming to the formal rules of grammar.     

Inferences,suppositions and explanatory extensions in argument interpretation

We describe a probabilistic approach for the interpretation of user arguments that integrates three aspects of an interpretation: inferences, suppositions and explanatory extensions. Inferences fill in information that connects the propositions in a user’s argument, suppositions postulate new information that is likely believed by the user and is necessary to make sense of his or her argument, and explanatory extensions postulate information the user may have implicitly considered when constructing his or her argument. Our system receives as input an argument entered through a web interface, and produces an interpretation in terms of its underlying knowledge representation—a Bayesian network. Our evaluations show that suppositions and explanatory extensions are necessary components of interpretations, and that users consider appropriate the suppositions and explanatory extensions postulated by our system. This article integrates and extends research described in George et al., 2004; Zukerman et al., 2004; Zukerman and George, 2005; George et al., 2005. The research described in this article was conducted while Sarah George was employed at Monash University and was supported in part by the ARC Centre for Perceptive and Intelligent Machines in Complex Environments.     

控制系统CAD软件包的设计——CSCAD的总体设计

本文提出一个CSCAD(Control System Computer Aided Design)软件包四级分块树状总体结构和管理命令的形式。叙述了CSCAD的命令系统和程序设计。这种结构和命令形式功能较强,便于用户依需要决定取舍和扩展。在计算机系统上实现也较为简单。同时便于移植。     

A command structured approach to structural object plotting and automated line dimensioning

The command structured DRAW1 system is a comprehensive plot program which allows the user to generate, position, display and dimension a complex object. Views or frames of the object in orthographic or perspective projections are automatically generated and displayed based upon commands denned by the engineer. The dimensioning of lines or points within a viewing frame is possible by using the standard criteria provided by the program, a user modification of that standard, or an original criteria selection defined by the user.Frame generation and base manipulation are effectively managed by a command interpreter and a flexible series of command forms. The command interpreter reads and interprets all information supplied to the system from a main line program. Real, integer, text and command form data are directed to a form interpreter which verifies the proper form and related element sequence. Free format command forms are provided to facilitate the manipulation of graphic data.Command and associated data are used to define, position and list all active coordinate and line information about a given object. Orthographic or perspective projections of an object are easily generated using pitch, yaw and roll commands to incrementally rotate the object into desired positions or orientations for frame plots.The automated dimensioning of a structured object may be accomplished by using the provided program standard criteria, which allows the user to select different types of arrow heads, dimensioning lines, architectural or engineering formats, stacking heights and dimensioning units. Any part of the standard criteria may be modified to better suit the user's needs; or, he may select his own special dimensioning criteria for his particular application. The user may define a standard dimensioning form to be used in dimensioning any line or line group.     

Modeling user action planning: A comprehension based approach

We review our efforts to model user command production in an attempt to characterize the knowledge users of computers have at various stages of learning. We modeled computer users with a system called NETWORK (Mannes and Kintsch, 1988; 1991) and modeled novice, intermediate, and expert UNIX command production data collected by Doane et al. (1990b) with a system called UNICOM (Doane et al., 1989a; 1991). We use the construction-integration theory of comprehension proposed by Kintsch (1988) as a framework for our analyses. By focusing on how instructions activate the knowledge rele/ant to the performance of the specified task, we have successfully modeled major aspects of correct user performance by incorporating in the model knowledge about individual commands and knowledge that allows the correct combination of elementary commands into complex, novel commands. Thus, experts can be modeled in both NETWORK and in UNICOM. We further show that salient aspects of novice and intermediate performance can be described by removing critical elements of knowledge from the expert UNICOM model. Results suggest that our comprehension-based approach has promise for understanding user interactions and implications for system design are discussed.Dr. Stephanie Doane is Assistant Professor of Psychology and appointed at the Beckman Institute at the University of Illinois. Shereceived her BAin Experimental Psychology from the University of California, Santa Barbara, her MS in Experimental Psychology from Villanova University, and her PhD in Cognitive Psychology from the University of California, Santa Barbara. Dr. Doane's research has focused on skill acquisition and the development and validation of theoretically-based computational models of cognitive processes. Her current research addresses issues of learning to interact with complex systems and the role of learning context in skill acquisition.Dr. Suzanne Mannes is Assistant Professor of Psychology at the University of Delaware. She received her BA in Psychology from the State University of New York College at Plattsburgh and received her PhD in Cognitive Psychology from the University of Colorado at Boulder. Her experimental research focuses on the role of prior knowledge in text comprehension, particularly as it pertains to problem-solving abilities. She also investigates the use of hybrid computer systems to simulate results from such studies.Dr. Walter Kintsch is Professor of Psychology and Director of the Institute of Cognitive Science at the University of Colorado in Boulder. He received his MA and PhD degrees in Experimental Psychology from the University of Kansas. His main area of interest has been the psychology of language and memory. He is currently the editor of the Psychological review.Peter Poison is Professor of Psychology and member of the Institute of Cognitive Science at the University of Colorado. He received his BA degree in Psychology and BS degree in Industrial Engineering from Stanford University and his PhD degree in Psychology from Indiana University. Dr. Poison's research has focused on the development and empirical evaluation of mathematical and computer simulation models of cognitive processes including transfer of training, problem solving, and the acquisition of cognitive skills. His current research deals with quantitative models of human-computer interaction and the application of such models to the design of more easily learned computer systems.     

ARINC661显控指令快速验证方法

针对ARINC661座舱显示系统中显示控制单元在开发阶段验证指令困难的问题,设计了一种解决方法.该方法不需要搭建联试环境,仅在控制指令开发的计算机上即可完成验证.该设计充分利用ARINC661通用内核进行扩展,在通用内核DF加载、指令解析、人机交互、画面渲染等基本功能的基础上,增加显示设备、虚拟UA的管理功能,既保证了辅助组件与实际显示画面的一致性,又能方便地模拟各种机载显示设备.该设计用直观的显示结果来验证控制指令的正确性,使设计人员能够方便快速地测试已开发指令的正确性,从而缩短试验周期,具有良好的适用性、扩展性和可靠性.     

Assisted navigation for a brain-actuated intelligent wheelchair

This paper presents an Assistive Navigation System (ANS) for a Robotic Wheelchair (RW) relying on a Brain–Computer Interface (BCI), as the Human–Machine Interface (HMI). A two-layer collaborative control approach is proposed to steer the RW, taking into account both user and machine commands. The first layer, a virtual-constraint layer, is responsible for enabling/disabling the user commands, based on context. More specifically, user commands are enabled for a set of situations requiring user decision, namely, bifurcations, multiple-directions caused by new obstacles in the environment, and deadlocks. The second layer is a user-intent matching responsible for determining the suitable steering command that better fits the user selection, taking into account the user competence to steer the wheelchair, and situation awareness of potential directions at a given location. A P300-based BCI allows the selection of commands to steer the RW. Experimental results using RobChair (Pires and Nunes (2002) [7], Lopes et al. (2007) [42]) are presented, showing the effectiveness of the proposed methodologies. The ANS was validated with ten able-bodied participants, and one participant with cerebral palsy, in two different scenarios: a structured known environment, and a structured unknown environment with moving objects. The overall result was that all participants were able to successfully operate the device, showing a high level of robustness of both, the BCI system, and the navigation system.     

Exploiting community feedback for information retrieval in DHT networks

In this paper, we propose CYBER, a CommunitY Based sEaRch engine, for information retrieval utilizing community feedback information in a DHT network. In CYBER, each user is associated with a set of user profiles that capture his/her interests. Likewise, a document is associated with a set of profiles—one for each indexed term. A document profile is updated by users who query on the term and consider the document as a relevant answer. Thus, the profile acts as a consolidation of users feedback from the same community, and reflects their interests. In this way, as one user finds a document to be relevant, another user in the same community issuing a similar query will benefit from the feedback provided by the earlier user. Hence, the search quality in terms of both precision and recall is improved. Moreover, we further improve the effectiveness of CYBER by introducing an index tuning technique. By choosing the indexing terms more carefully, community-based relevance feedback is utilized in both building/refining indices and re-evaluating queries. We first propose a naive scheme, CYBER+, which involves an index tuning technique based on past queries only, and then re-evaluates queries in a separate step. We then propose a more complex scheme, CYBER+ +, which refines its index based on both past queries and relevance feedback. As the index is built with more selective and accurate terms, the search performance is further improved. We conduct a comprehensive experimental study and the results show the effectiveness of our schemes.     

The design of an interactive graphics system for the display of meteorological fields

We describe a computer graphics system for the display of meteorological fields (such as plots of temperature and pressure). Much of the design effort has gone into the command processor which uses the usual command format of VERB PARAMETER = VALUE … The parameter and values are defined in terms of a tree sructure and the program searches as much of the tree as is necessary to make an unambiguous indentification of a token (where such an identification is possible). Because of this, commands can generally be abbreviated to the format VERB VALUE. This shortened form is possible even when the value of the command parameter is numerical. This is achieved by always specifying the units of numerical values (such as 5 days, 24 hours, 1000 millibars, etc.). The command format allows the user to manipulate a number of fields to form derived fields such as the difference of two pressure forecasts. The use of defaults for all parameters (which change as the user requests plots), reduces the amount of typing required to a minimum. A ‘Help’ facility aids the user in his choice of commands.     

Electromyography-based control of active above-knee prostheses

This paper presents a new electromyography (EMG)-based control approach for above-knee (AK) prostheses, which enables the user to control the prosthesis motion directly with his or her muscle activating neural signals. Furthermore, the unique ‘active-reactive’ control structure mimics the actuation mechanism of a human biological joint, and thus provides the user an experience similar to that of a biological lower limb in the control process. In the proposed control approach, surface EMG is utilized to provide a non-intrusive interface to the user's central nervous system, through which the muscle-activating signals can be obtained. With the EMG signals as inputs, an ‘active-reactive’ control algorithm is developed based on the analysis on a simplified musculoskeletal structure of human biological joint. This control algorithm incorporates an ‘active’ component, which reflects the user's active effort to actuate the joint, and a ‘reactive’ component, which models the reaction of the joint to the motion as a result of the controllable impedance displayed on the joint. With this unique structure, the controller enables the active control of the joint motion, while at the same time achieves a natural interaction with the environment through the modulation of the joint impedance. The effectiveness of the proposed control approach was demonstrated through a set of free swing experiments, in which the user was able to control the prosthesis to follow arbitrary motion commands, and a set of level walking experiments, in which the user achieved natural walking gait similar to the typical walking gait of healthy subjects.     

Look where you're going [robotic wheelchair]

We propose a robotic wheelchair that observes the user and the environment. It can understand the user's intentions from his/her behaviors and the environmental information. It also observes the user when he/she is off the wheelchair, recognizing the user's commands indicated by hand gestures. Experimental results show our approach to be promising. Although the current system uses face direction, for people who find it difficult to move their faces, it can be modified to use the movements of the mouth, eyes, or any other body parts that they can move. Since such movements are generally noisy, the integration of observing the user and the environment will be effective in understanding the real intentions of the user and will be a useful technique for better human interfaces.     

The role of excursions in interactive systems

It is recommended that man—machine dialogue systems be designed so as to make them more self-explanatory, and hence more user-friendly, by the inclusion of a dialogue form called an excursion tour. This is defined as an information-gathering sequence of operations that enables the user to learn the commands which change his data sets, and which are called direct way commands. The distinction between the excursion and the direct way forms of dialogue is justified in terms of cognitive psychological theory, and corresponds to the distinction made there between planning and performance, or between knowing and doing. The operation of the two dialogue forms and their interaction is represented by an extension of the TOTE model of human task performance. The term interactive deadlock, or unsolvable incongruence, is introduced to denote the situation in which either no command exists that will further an essential subgoal, or if it does exist the user has no way of discovering it within the system. In terms of the TOTE model, it is shown how such deadlocks or incongruences that are unsolvable on one level may be overcome by extending the user's knowledge with the help of excursions.     

SUN工作站用户界面的改进

本文以ＳＵＮ工作站为背景，提出了ＮＩＳＸ操作系统用户界面的改进方法，改进后的用户面对初学者是友好的，用户可以根据清晰的命令格式说明和上下文有关的帮助信息。直接操纵ＵＮＩＸ的各种命令。程序设计的基本特点是采用了命令表数据结构和ＳｕｎＶｉｅｗ工具，从而使编程非常容易而且简洁。     

Host-based intrusion detection using dynamic and static behavioral models

Intrusion detection has emerged as an important approach to network security. In this paper, we adopt an anomaly detection approach by detecting possible intrusions based on program or user profiles built from normal usage data. In particular, program profiles based on Unix system calls and user profiles based on Unix shell commands are modeled using two different types of behavioral models for data mining. The dynamic modeling approach is based on hidden Markov models (HMM) and the principle of maximum likelihood, while the static modeling approach is based on event occurrence frequency distributions and the principle of minimum cross entropy. The novelty detection approach is adopted to estimate the model parameters using normal training data only, as opposed to the classification approach which has to use both normal and intrusion data for training. To determine whether or not a certain behavior is similar enough to the normal model and hence should be classified as normal, we use a scheme that can be justified from the perspective of hypothesis testing. Our experimental results show that the dynamic modeling approach is better than the static modeling approach for the system call datasets, while the dynamic modeling approach is worse for the shell command datasets. Moreover, the static modeling approach is similar in performance to instance-based learning reported previously by others for the same shell command database but with much higher computational and storage requirements than our method.