ServicenavigationHauptnavigationTrailKarteikarten


Research unit
COST
Project number
C00.0105
Project title
Image and speech processing for personalized multimedia communication terminals

Texts for this project

 GermanFrenchItalianEnglish
Key words
-
-
-
Anzeigen
Research programs
-
-
-
Anzeigen
Short description
-
-
-
Anzeigen
Further information
-
-
-
Anzeigen
Partners and International Organizations
-
-
-
Anzeigen
Abstract
-
-
-
Anzeigen
References in databases
-
-
-
Anzeigen

Inserted texts


CategoryText
Key words
(English)
Image processing in the compressed domain; information management in the compressed domain; image and low bit-rate video compression; wideband speech coding; spectral quantization for wideband speech coding; voice interface for portable devices.
Research programs
(English)
COST-Action 276 - Information and knowledge management for integrated media communication systems
Short description
(English)
See abstract
Further information
(English)
Full name of research-institution/enterprise: Université de Neuchâtel Institut de Microtechnique Laboratoire d'Electronique et de Traitement du signal
Partners and International Organizations
(English)
AT, BG, HR, CY, CZ, MK, FR, EL, HU, IE, IT, NO, RO, CS, SK, SI, ES, CH, TR
Abstract
(English)
Originally foreseen for three years (01.07.2001 - 30.06.2004), the project has been extended by one year - without modification of the allocated budget - in view of completing the research and publishing the results in image processing and filtering. Voice processing was not handled over the fourth year of the project. The following topics were covered during the last year: Image processing in the compressed domain: Today most digital images and video sequences are stored and transmitted in compressed form. This generalized compressed status of visual information in multimedia systems and networks promotes a large interest in the research and development of image and video processing algorithms that operate directly in the so-called compressed- or frequency-domain. The advantage of these compressed-domain algorithms is that they avoid the computationally-intensive decompression stage. An additional and important reduction in the computational complexity is also achieved by the fact that the number of data to process is significantly less in the compressed domain with respect to the pixel domain algorithms. It is important to point out that this drastic reduction in the computational complexity and consequently in power consumption can be obtained without compromising the robustness and efficiency of the algorithms are reported for several studied applications in the sections below. Polynomial framework for designing digital and classical filters: A framework for designing a wide range of various filter classes fulfilling amplitude and/or phase response specifications was elaborated over several years between the Budapest University of Economics & Technology and Uni-Neuchâtel. Purpose of the cooperation within COST 276 was to assemble and work out the results for a comprehensive publication. Further activities: The research cooperation established with national and international projects enabled the presentation of several joint conference publications, in particular in the context of an STSM with the Technical University of Iasi / Romania. Moreover, the knowledge acquired gave rise to two colloquia presented in the frame of a regular course at Uni-Neuchâtel, whereas a one-week intense course on advanced signal processing co-organized with TU-Iasi was delivered in Iasi / Romania in May 2005 with support from COST 276 and SCOPES / Switzerland.
References in databases
(English)
Swiss Database: COST-DB of the State Secretariat for Education and Research Hallwylstrasse 4 CH-3003 Berne, Switzerland Tel. +41 31 322 74 82 Swiss Project-Number: C00.0105