Maya Dimitrova

NEURAL NETWORKS FOR AIR POLLUTION NOWCASTING

This work illustrates the use and some results of artificial neural networks (ANNs) for data quality control of air pollutants. ANNs are applied to the short-term predicting of air pollutant concentrations in urban areas. Observed versus predicted data are compared to test the efficacy of ANNs in simulating environmental processes. Statistical analysis is used for choice of neural structure. The model is validated on original data.

Designing Humanoid Robots with Novel Roles and Social Abilities

This mini-review describes aspects of human-robot interaction to be taken into account when designing humanoid robots with novel roles and social abilities. The review accentuates the psychological complexity that is necessary to be made inherent in the design of humanoid robotic technology. Some recent studies of robot acceptance are summarized leading to the proposal for more complex synthetic sensors being needed in novel humanoid robots. The perspective is designing based on modeling attitude (the social level of human robot interaction), but not opinion (psychological level), which can be a valuable aim for humanoid robotics.

Personality Filter in Mobile Networks with Communication Constraints

A new psychological model for efficient data transmission in mobile networks under communication constrains is proposed. It accounts for user personality characteristics to determine the feasible path for packet transmission from source to target and to predict the availability of the communication resources based on dynamically determined level of node generosity. Two case studies are presented with initial performance better than chance. The model is designed for the developed framework of evolving fuzzy modeling in mobile Ad hoc networks via lightweight online unsupervised learning.

Designing a system of interactive robots for training collaborative skills to autistic children

Using robots to reward and stimulate children doing tasks together can be helpful in improving their social skills. Robots implemented 3 types of behavior scenarios - imitate, enhance and counteract - in a collaborative game of teaching robots perform movements shown by hand gestures. The paper presents a novel account of ASD symptoms based on the interplay of empathic concern and empathic accuracy in different personalities with and without cognitive deficits. Performance of autistic children was compared with two age groups of typically developing children - under and above the age of 7. The study confirmed the limited resource hypothesis - cognitive and motivational - bringing about the observed phenomena in situations of limited attention resources and performance under distraction. The robotic framework for educating children has shown potential for developing of various complex cognitive and social skills and has substantial developmental impact.

A framework for design of cloud compatible medical interfaces

A framework for design of novel interfaces for Cloud compatible medical services called User Interface as a Service (UIaaS) is presented in the paper. The aim is to support seamless and ubiquitous health monitoring based on flexible and useful Cloud services for healthcare. The paper describes the methods and new interfaces for health monitoring that have recently been developed for the proposed framework. The influence of the new methods and technologies for future Cloud computing systems for healthcare is discussed.

Hand gestures recognition based on lightweight evolving fuzzy clustering method

Robots of the future will socialize with humans. Human-robot interaction (HRI) by a vision-based gesture interface helps to personalize the communication with humans in various contexts - from support of their daily life to social skills training of children with developmental problems. We are especially interested in vision-based hand gesture HRI and propose a hand gesture recognition system based on a novel online extraction and classification scheme, which is lightweight and can be used in a mobile robot. An online Lightweight Evolving Fuzzy Clustering Method is used to categorize the positional and HSV model of pixels for the edges of the gesture image. The result clusters consist of (x, y) coordinates and the averaged grayscale level at these locations. Then these clusters are processed to identify typical for the hand features brighter and darker pixel information. The database consists of averaged grayscale levels in HSV format for neighbor pixels that characterize different features. For feature recognition we use Tanimoto similarity measure for matching the current grayscale patterns to those in the database. Then the feature location is encoded in a binary format. For gesture recognition we use a formalism of Symbol Relation Grammars to describe a gesture, as well as simple and fast bitwise operations to find the position and orientation of the features in the gesture.

Socially-Competent Computing Implementing Social Sensor Design

The paper presents a conceptual model for social sensor design in socially-competent computing systems. The model is based on theories of social behavior being driven by the underlying attitudes, rather than on models predicting behavior in response to behavior representing people as physical objects in dynamic interactions. It is proposed to increase the ability of the systems to extract relevant features and to achieve better social competence, similar to the kind that is underlying human interactions by implementing algorithms, capable of predicting behavior in response to attitude. The paper presents an account of the social level of understanding human interactions in the context of three application scenarios-multi-hop communication networks, embedded systems for support of medical interventions and information systems supporting educational activities. Patterns of real data are discussed in terms of the proposed model of social sensor design for enhanced socially-competent computing.

Integration of multiple sensor modalities in active vessel cardiology workstation

Active vessel is a new multimedia workstation which enables the visualization, acquisition and handling of different medical image modalities on- and offline for cardiology purposes. It implements DICOM v3.0 decompression and browsing, video acquisition, reproduction and storage for intravascular ultrasound (IVUS) and angiogram analysis. A distinctive implementation feature is the automatic catheter segmentation in angiography images. Another distinctive feature is the interactive mode of model correction via mouse dragging. A third and novel technical solution of the catheter path reconstruction is the integration with a magneto-sensitive micro-device inside the catheter for faster and safer minimal invasive surgical intervention. The paper gives the overview of the entire system, its basic new functionalities and the proposed technical solution of integration of four (instead of three) sensor modalities in cardiovascular practice.

A Method to Evaluate Intuitive Sense by Using a Robotic Tool: Towards Engineering for Assistive Technology and Accessibility

The evaluation of subjective feelings of discomfort is a serious problem in the field of ergonomics, environment engineering and human-robot interactions. It is difficult to measure them by using objective quantifications in a traditional sense. We have hypothesized that the unpleasant feeling emerges caused by deviating from the autonomy, which exists in the perception-action cycle, the convective flow in two layers of intuition and perception and continuity in the dialogue. The present paper discusses the issue with respect to two examples; an unexpected breaking of balance in the trained motion and irritation at the lack of progress by discontinuity in the dialogue, and attempts to build a theoretical framework to detect a deviation from ongoing autonomy. This paper discusses a way of the theoretical modeling and explores possible implementations into the robotic experiment that reproduces human behaviors in need of supervision by the central nervous system. This preliminary report can contribute to an extension of the coupling between the central and peripheral nervous systems to social communications which is supported by assistive technologies and accessibility improvements.

Adaptive and intuitive interactions with socially-competent pedagogical assistant robots

The paper presents a novel framework for including social competence in pedagogical assistant robots based on recent social neuroscience investigations. An educational model describing the interplay of social interaction and social observation in typical children and children with autistic spectrum conditions (ASC) is proposed. The behavioral results of a study of human gaze movements while viewing a robot tutor are briefly presented. The implications for design of adaptive and intuitive robotic tutoring systems as pedagogical assistants for typical children and children with special educational needs are discussed.