Abstract

The growth of cellular telephony combined with recent advances in speech recognition technology results in sizeable potential opportunities for mobile speech recognition applications. Classic robustness techniques that have been previously proposed for speech recognition yield limited improvements of the degradation introduced by idiosyncrasies of the mobile networks. These sources of degradation include distortion introduced by the speech codec as well as artifacts arising from channel errors and discontinuous transmission. In this thesis we focus on characterizing the distortion introduced to the speech signal by the speech codec and we propose methods for reducing the detrimental effect of coding on recognition accuracy. The initial focus of this thesis is on the full rate GSM codec (FRGSM) . We propose a method to generate recognition features directly from codec parameters. It is shown in this work that by selectively constructing a cepstral feature vector from the GSM codec para...

Citations

1180	Fundamentals of speech recognition - Rabiner - 1993
874	Error bounds for convolutional codes and an asymptotically optimum decoding algorithm - Viterbi - 1967
612	Statistical Methods for Speech Recognition - Jelinek - 1997
550	Comparison of parametric representations of monosyllabic word recognition in continuously spoken sentences - Davis, Mermelstein - 1980
443	Perceptual linear predictive (PLP) analysis of speech - Hermansky - 1990
440	Maximum likelihood linear regression for speaker adaptation of continuous density hidden markov models - Leggetter, Woodland - 1995
345	A maximum likelihood approach to continuous speech recognition - Bahl, Jelinek - 1983
178	Cepstral analysis technique for automatic speaker verification - Furui - 1981
156	Some statistical issues in the comparison of speech recognition algorithms - Gillick, Cox - 1989
145	Environmental robustness in automatic speech recognition - Acero, Stern - 1990
133	Effectiveness of linear prediction characteristics of the speech wave for automatic speaker identification and verification - Atal - 1974
113	Multi–stream speech recognition - Bourlard, Dupont, et al. - 1996
82	Line Spectrum Representation of Linear Predictive Coefficients of Speech Signals - Itakura - 1975
78	Robust continuous speech recognition using parallel model combination - Gales, Young - 1996
72	Speech recognition in noisy environments - Moreno - 1996
58	The Harpy Speech Understanding System - Lowerre, Reddy - 1980
47	Discriminative Model Combination - Beyerlein - 1998
43	Design of a linguistic statistical decoder for the recognition of continuous speech - Jelinek, Bahl, et al. - 1975
35	Subphonetic Acoustic Modeling for Speaker-Independent Continuous Speech Recognition - Hwang - 1993
34	Discrete time signal processing - Oppenheim, Schaffer, et al. - 1999
24	Quantization of cepstral parameters for speech recognition over the world wide web - Digalakis, Neumeyer - 1999
22	Compression of acoustic features for speech recognition in network environments - Ramaswamy, Gopalakrishnan - 1998
21	The DoD 4.8 kbps Standard (Proposed Federal Standard 1016 - Jr, Tremain, et al. - 1991
18	The Mobile Phone Meets the Internet - Oliphant - 1999
17	Channel normalization techniques for automatic speech recognition over the telephone - Veth, Boves - 1998
17	Deleted interpolation and density sharing for continuous hidden Markov models - Huang, Hwang, et al. - 1996
13	Towards subband-based speech recognition - Bourlard, Dupont, et al. - 1996
13	Regular-pulse excitation } A novel approach to e!ective and e$cient multipulse coding of speech - Kroon, Deprettere, et al. - 1986
12	Wireless Access and the Local Telephone Network - Calhoun - 1992
12	The in¯uence of speech coding algorithms on automatic speech recognition - Euler, Zinke - 1994
8	Robust speech recognition for multiple topological scenarios of the GSM mobile phone system - Salonidis, Digalakis - 1998
7	Deconvolution of Telephone Line Effects for Speech Recognition - Mokbel, Jouvet, et al. - 1996
6	Introduction to Digital Mobile Communication - Akaiwa - 1997
6	Laid NM, Rubin DB: Maximum likelihood from incomplete data via the EM algorithm - AP - 1977
3	Speech recognition for mobile communications - Haavisto - 1999
3	Robust Speech Recognition for Wireless Networks and Mobile Telephony - Haeb-Umback - 1997
3	Distortion-class weighted acoustic modeling for robust speech recognition under GSM RPE-LTP coding - Huerta, Stern - 1999
3	An evaluation of 4800bps voice coders - Kemp, Sueda, et al. - 1989
3	Cellular phone speech recognition: noise compensation vs. robust architectures - Puel, Andre-Obrecht - 1997
3	Adapting PSN recognition models to the GSM environment by using spectral transformation - Soulas, Mokbel, et al. - 1997
2	GSM speech compression software implementation. ftp://ftp.cs.tu-berline.de/pub/local/ kbs/tubmik/gsm - Degener, Bormann - 1992
2	Frame-synchronous adaptation of cepstrum by linear Regression - Delphin-Poulat, Mokbel - 1997
2	Name dialing using final user defined vocabularies - Elvira, Torrecilla - 1998
2	Paliwal K.K, “An introduction to Speech coding - Kleijn - 1995
2	Improving speech recognition performance by using multimodel approaches - Ming, Hanna, et al. - 1999
2	Automatic detection of disturbing robot voice and ping pong e€ects in GSM transmitted speech - Paping, Fahnle - 1997
1	Speech Related features in GSM”, in: GSM: Evolution Towards 3rd Generation Systems. Edited by Zvonar - Aftelak, “New - 1999
1	Evaluation of the Root-Normalised Front-end (RM_LFCC) for SPeech Recognition - Dufour, Glorion, et al. - 1996
1	Speech Recognition over GSM: SPecific features and Performance Evaluation - Fissore, Ravera, et al. - 1999
1	Diaz-de-Maria F., Valverde-Albacete F., “Avoiding Distortions due to Speech Coding and Transmission - Gallardo-Antolin - 1999