Karla Conn Welch

An Approach to the Design of Socially Acceptable Robots for Children with Autism Spectrum Disorders

Investigation into technology-assisted intervention for children with autism spectrum disorders (ASD) has gained momentum in recent years. Research suggests that robots could be a viable means to impart skills to this population since children with ASD tend to be fascinated by robots. However, if robots are to be used to impart social skills, a primary deficit for this population, considerable attention needs to be paid to aspects of social acceptability of such robots. Currently there are no design guidelines as to how to develop socially acceptable robots to be used for intervention for children with ASD. As a first step, this work investigates social design of virtual robots for children with ASD. In this paper we describe the design of a virtual environment system for social interaction (VESSI). The design is evaluated through an innovative experiment plan that combines subjective ratings from a clinical observer with physiological responses indicative of affective states from the participants, both collected when participants engage in social tasks with the social robots in a virtual reality environment. Two social parameters of importance for this population, namely eye gaze and social distance, are systematically varied to analyze the response of the participants. The results are presented to illustrate how experiments with virtual social robots can contribute towards the development of future social robots for children with ASD.

An Affect-Sensitive Social Interaction Paradigm Utilizing Virtual Reality Environments for Autism Intervention

This paper describes the design and development of both software to create social interaction modules on a virtual reality (VR) platform and individualized affective models for affect recognition of children with autism spectrum disorders (ASD), which includes developing tasks for affect elicitation and using machine-learning mathematical tools for reliable affect recognition. A VR system will be formulated that can present realistic social communication tasks to the children with ASD and can monitor their affective response using physiological signals, such as cardiovascular activities including electrocardiogram, impedance cardiogram, photoplethysmogram, and phonocardiogram; electrodermal activities including tonic and phasic responses from galvanic skin response; electromyogram activities from corrugator supercilii, zygomaticus major, and upper trapezius muscles; and peripheral temperature. This affect-sensitive system will be capable of systematically manipulating aspects of social communication to more fully understand its salient components for children with ASD.

Physiological signals of autistic children can be useful

This article covers the latest research concerning the measurement of physiological signals of children with autism, particularly for the study of changing emotions in various environments. Answers to important questions regarding autistic childrens physiological activity are examined, and we will see that within a non-social environment, physiological responses are the same between children with and without autism but different in environments with social contexts. Moreover, physiological signals can be used as a reliable indicator of emotions of children with autism. Also covered are the latest developments in wearable sensor technologies avail- able for measuring on-the-go. I review additional research that identifies body signals in response to stimuli and may help ex- plain core social deficits in children with autism.

A review of electricity monitoring and feedback systems

This paper reviews various products available in market and under study, which monitor electricity consumption and provide feedback of the energy usage. Energy usage feedback systems are categorized into three groups: (i) socket monitoring systems, (ii) whole-house monitoring systems, and (iii) whole-house monitoring with breakdown to individual appliances. In the future directions section, the paper describes a proposed energy usage feedback system, which provides feedback about individual appliances. This system also considers the effect of human behavior on energy usage and suggests a possible solution.

EMF Signature for Appliance Classification

Various intrusive and nonintrusive appliance load monitoring and classification systems have been studied; however, most of them designed so far provide group-level energy usage feedback. We present the first phase of a system with the potential to attribute energy-related events to an individual occupant of a space and provide occupant-specific energy usage feedback in an uninstrumented space (e.g., home or office). This initial phase focuses on collecting the electromagnetic field (EMF) radiated by several common appliances to determine a unique signature for each appliance. It also implements a machine learning algorithm to classify appliances from an incoming EMF data file. The proposed approach has been prototyped with hardware realization. The results obtained on tested appliances indicate the EMF sensors ability and potential to develop a system for providing occupant-specific energy feedback.

Using Robot-Assisted Instruction to Teach Students with Intellectual Disabilities to Use Personal Narrative in Text Messages

In the current investigation, we evaluated the effectiveness of a multi-component package (i.e., robot, simultaneous prompting, self-graphing) for teaching three students, ages 19–21, with intellectual disabilities (ID) to write text messages that included a greeting, personal narrative, and closing. Data suggest that the package was effective in increasing correct performance for all participants. In addition, participants demonstrated their newly acquired texting skills across different communicative partners.

A new shape-based framework for the left ventricle wall segmentation from cardiac first-pass perfusion mri

We propose a shape-based approach for the segmentation of the left ventricle (LV) wall on cardiac first-pass magnetic resonance imaging (FP-MRI) using level sets. To reduce the variabilities of the LV wall in FP-MRI, it is first imperative to co-align the time series images to account for the global and local motions of the heart. Therefore, we developed a two-step registration methodology that includes an affine-based registration followed by a local B-splines based alignment to maximize a similarity function that accounts for the first- and second-order normalized mutual information (NMI). Additionally, myocardial signal intensity varies with the agent transit, which makes it difficult to control the level set evolution using image intensities alone. Thus, we constrained the level set evolution using three features: a weighted probabilistic shape prior, the first-order pixel-wise image intensities, and a second-order Markov-Gibbs random field (MGRF) spatial interaction model. We tested our approach on 24 data sets in 8 infarction patients using the Dice similarity coefficient (DSC), comparing our approach to other shape-based segmentation approaches. We also tested the performance of our segmentation approach using the receiver operating characteristics (ROC). Our approach achieved a mean DSC value of 0.910±0.037 compared to other shape-based methods that achieved 0.862±0.045 and 0.844±0.047. Finally, the ROC analysis for our segmentation method showed the best performance, with area under the ROC curve of 0.92, while that for intensity showed the worst performance, with area under the ROC curve of 0.69.

Understanding psychophysiological response to a Virtual Reality-based social communication system for children with ASD

Deficits in social communication skills are thought to be one of the core deficits in children with autism spectrum disorders. Specifically, these children are characterized by communicative impairments, particularly regarding expression of affective states. However, they often experience states of emotional or cognitive stress measured as Autonomic Nervous System activation without proper external expression placing limits on traditional conversational and observational methodologies. In recent years, several assistive technologies, particularly Virtual Reality (VR), have been investigated to promote social interactions in this population. Here we present the development of a VR-based social communication system that is made affect-sensitive by using a physiology based approach.

Modeling Human Behavior for Energy-Usage Prediction

We propose a system that uses a set of mobile sensors to model human behavior of energy usage. This mobile sensor suite can be fit on a keychain or ID/access badge. Data from these sensors, e.g., temperature, visible light spectrum, and 60 Hz electromagnetic field, will be used to give real-time feedback of user’s energy consumption and prediction of future energy usage. Feedback of energy consumption will be displayed in an understandable manner on a user interface, e.g., smart phone. A model developed from the available data using machine learning will inform the system about energy consumption patterns and behaviors of users.

The Use of an Autonomous Pedagogical Agent and Automatic Speech Recognition for Teaching Sight Words to Students With Autism Spectrum Disorder

In the current study, we examined the effects of an instructional package comprised of an autonomous pedagogical agent, automatic speech recognition, and constant time delay during the instruction of reading sight words aloud to young adults with autism spectrum disorder. We used a concurrent multiple baseline across participants design to evaluate the efficacy of intervention and conducted post-treatment probes to assess maintenance and generalization. Our findings suggest that all three participants acquired and maintained new sight words and demonstrated generalized responding.