Professor A. Ariyaeeinia
BEng MSc PhD CEng MIET
Job title: Professor of Signal Processing
Email Address: A.M.Ariyaeeinia@herts.ac.uk
Memberships and Appointments: Editorial Board member, IET Signal Processing Journal Chairman, Management Committee, COST Action 275 Member, Steering Committee, IRIS (EPSRC Network of Academics in Biometrics) Member, Advisory Board, UK Biometrics Institute Member, Management Committee, COST Action 2101
Ariyaeeinia is a Professor of Signal Processing in the Faculty of Science, Technology and Creative Arts. Over the last eighteen years, he has been actively involved in research in close collaboration with industry and has published numerous papers. The focus of his research during this period has included voice biometrics (speaker identification/verification), speaker change detection in audio documents and multimodal biometric fusion. As a result of the research activities in the above areas, he founded the Speech & Biometrics lab in 1998 which has since become a valuable basis for collaboration with UK industry and for research cooperation with European partners. Ariyaeeinia is a member of the Editorial board of the IET Signal Processing Journal. He has served on the scientific committees of various international conferences, and is an active referee for a number of academic journals.
Ariyaeeinia's research efforts over the past few years have included a broad range of activities including various national and European projects. He was a co-proposer and is now a steering committee member of an EPSRC Network of Academics in Biometrics. Ariyaeeinia is a founding partner and also a member of the Advisory Board of the UK Biometrics Institute. Between 2001 and 2005, he led EC-COST 275 (Biometrics on the Internet) involving thirty research establishments from fourteen European countries. In 2003, Ariyaeeinia acted as the Guest Editor for the IEE Proc-VIS first Special Issue on Biometrics. In 2006, he was appointed by the Office of Science and Innovation to represent the UK on the Management Committee of COST 2101 (Biometric Identification: 2006-2010). His research into speaker change detection/indexing attracted the collaboration of 20/20 Speech and Canon Research Lab, and led to a breakthrough in the field in 2007. Ariyaeeinia's research collaboration with BT in speaker recognition resulted in a number of patents, and led to the current collaborative project with BT Security Research Group in the field of multimodal biometrics.
Research Interests
Speaker and language recognition, speech processing and enhancement, speaker-based audio-visual data indexation, biometrics-based recognition over the Internet, multimodal biometrics, stereoscopic imagingTeaching specialisms
Digital Signal Processing, Speech Processing, Biometrics, Signals & SystemsCollaborations and Projects
Multimodal biometrics: 2004-present The aim of this project, which is conducted in collaboration with BT Security Research Group, is to improve the identification/verification accuracy beyond what is achievable with a single type of biometrics. An important aspect of the work in this area is that of minimising the effects of degradation in biometric data. Such a capability can be of considerable value in enhancing the recognition accuracy through multimodal biometrics when operating in an uncontrolled environment.
Biometrics for identity documents and smart cards (BIDS):2006-2010 The main objective of this COST project is to investigate effective biometric authentication techniques for deployment in the next generation of biometrics-enabled identity documents and smart cards. The work is conducted in collaboration with researchers from 14 other EU countries.
Speaker-based indexation of conversational audio: 2000-Present This project is concerned with the development of effective methods for indexing conversational audio based on voice biometrics. The process involves partitioning audio into speaker homogeneous segments and then labelling each segment with the identity of the speaker. Such a process has various potential applications including automatic management of diverse collections of radio/TV broadcast material. The project has benefited from several years of collaboration with industry.
User voice identification (URVIN):2003-2006 This project was part of the DTI Next Wave Technologies and Markets programme. It was conducted in collaboration with Fulcrum Voice Technologies and in association with the DTI Virtual Research Centre at the University of York. The work was mainly concerned with the development of a system for the extraction of robust speech features in uncontrolled environment. The project led to the development of an integrated noise cancellation-feature extraction system and a demonstrator for a user classification scenario.
Biometrics-based recognition over the Internet: 2001-2005 The main objective of the project, which was conducted within the COST framework, was to investigate effective methods for the recognition of people over the Internet based primarily on their voice and facial characteristics. The motivation behind this research was the need for a reliable means of identification to complement/substitute the traditional identity verification approaches deployed for logical access to services over the Internet.