Akbar Ghobakhlou

An adaptive model of person identification combining speech and image information

The paper introduces a combination of adaptive neural network systems and statistical method for integrating speech and face image information for person identification. The method allows for the development of models of persons and their on-going adjustment based on new speech and face images. The method is illustrated with a modeling and classification of different persons, when speech and face images are presented in an incremental way. In this model, there are two subnetworks, one for face image and one for speaker recognition. A higher-level layer is applied to make a final decision. In the speaker recognition subnetwork, a text-dependant model is built using evolving connectionist systems (ECOS) [N. Kasabov, 2002]. In the face image recognition subnetwork, composite profile technique is applied for face image feature extraction and zero instruction set computing (ZISC) [ZISC Manual, 2000] technology is used to build the neural network. In the higher-level conceptual subsystem, final recognition decision is made using statistical method. The experiments show that ECOS and ZISC are appropriate techniques for the creation of evolving models for the task of speaker and face recognition individually. It is also shown that the integration of the speech and image information using statistical method improves the person identification rate.

A service oriented wireless sensor node management system

Wireless Sensor Network (WSN) is rapidly evolving technology and has been applied in many fields, such as tracking and environment monitoring applications. Exponential growth in sensor data availability demands efficient data management and sharing. Thus integrated WSN with web services are becoming common in widespread applications across the world. There is a high demand for standardising access to sensor data via internet without having to use some complex and unknown protocol. Service Oriented Architecture (SOA) is one of the key paradigms that enables the deployment of services at large-scale over the internet domain and its integration with WSNs open new pathways for novel applications and research. This paper presents a SOA for WSN node management and describes an implementation of services designed for crop monitoring application.

An Evolving Neural Network Model for Person Verification Combining Speech and Image

This paper introduces a method based on Evolving Connectionist Systems (ECOS) for person verification tasks. The method allows for the development of models of persons and their on-going adjustment based on new speech and face images. Some experimental person verification models based on speech and face image features are developed based on this method where speech and face image information are integrated at a feature level to model each person. It is shown that the integration of speech and image features improves significantly the accuracy of the person verification model when compared with the use of only image or speech data.

In-vehicle Noise and Enhanced Speech Intelligibility

In-Car speech recognition will be pervasive over the coming years. The goal of speech enhancement is to increase the quality and intelligibility of speech in a noisy environment. The focus of the present research is to evaluate the effect of speech enhancement on the intelligibility of spoken language in a moving vehicle. Here, an ECoS network is used as a model to evaluate the intelligibility. A baseline performance was established using clean speech data. This data was then mixed with various types of in-vehicle noise at several signal-to-noise ratios. Speech enhancement techniques were applied to the noisy speech data. The performance of the ECoS model was evaluated when the noisy and enhanced speech was presented. Several factors were found to affect the recognition rate, including noise type and noise volume.

Implementing and Evaluating Snapshots + Events Spatiotemporal Modelling Approach

Traditional approaches for modelling spatiotemporal information (snapshots, states and events) are not very efficient and usually not capable of retrieving information based on specific spatiotemporal query. The snapshots modelling technique is the oldest and simplest approach and does not support any spatiotemporal query. The most current approach is the events modelling technique which allows data retrieval using spatiotemporal query. However, deductive capacities are needed for developing a system which is not usually present in traditional databases. One idea is to retain the advantages of each individual traditional approach while combining them to build and develop an efficient spatiotemporal information system. A case study is presented for building a spatiotemporal information system combining snapshots and events approaches which overcomes some of the problems associated with snapshots and events approaches on their own. This work examine the hybrid (snapshots + events) spatiotemporal modeling technique with generic implementation details.