Zobrazeno 1 - 10
of 12
pro vyhledávání: '"Antti Hurmalainen"'
Autor:
Martin Wöllmer, Gerhard Rigoll, Jort F. Gemmeke, Tuomas Virtanen, Björn Schuller, Jürgen T. Geiger, Antti Hurmalainen, Felix Weninger
Publikováno v:
ICASSP
This paper proposes a multi-stream speech recognition system that combines information from three complementary analysis methods in order to improve automatic speech recognition in highly noisy and reverberant environments, as featured in the 2011 PA
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::ff2ccb4b2a18ef57fb6ceae464f00612
https://opus.bibliothek.uni-augsburg.de/opus4/frontdoor/index/index/docId/71522
https://opus.bibliothek.uni-augsburg.de/opus4/frontdoor/index/index/docId/71522
In this study, we propose an unsupervised method for dictionary learning in audio signals. The new method, called binary nonnegative matrix deconvolution (BNMD), is developed and used to discover patterns from magnitude-scale spectrograms. The BNMD m
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::3177af42532a6c34dc0eced93df8e538
https://trepo.tuni.fi/handle/10024/126250
https://trepo.tuni.fi/handle/10024/126250
Publikováno v:
Computer Speech & Language. 27:763-779
Speech recognition systems intended for everyday use must be able to cope with a large variety of noise types and levels, including highly non-stationary multi-source mixtures. This study applies spectral factorisation algorithms and long temporal co
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::31b34f6c34a4e4ffb46856fb425ff504
https://doi.org/10.1016/j.csl.2012.07.008
https://doi.org/10.1016/j.csl.2012.07.008
Publikováno v:
EUSIPCO
Sparse representations have been found to provide high classification accuracy in many fields. Their drawback is the high computational load. In this work, we propose a novel cascaded classifier structure to speed up the decision process while utiliz
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::b6b1880989f5e44375c59c015687d34f
https://doi.org/10.1109/eusipco.2016.7760660
https://doi.org/10.1109/eusipco.2016.7760660
Publikováno v:
INTERSPEECH
Recognition and classification of speech content in everyday environments is challenging due to the large diversity of realworld noise sources, which may also include competing speech. At signal-to-noise ratios below 0 dB, a majority of features may
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::b17ff306b1ecc8ec72aa7dc62d4cf935
https://doi.org/10.21437/interspeech.2015-93
https://doi.org/10.21437/interspeech.2015-93
Publikováno v:
ICASSP
Non-negative matrix factorisations are used in several branches of signal processing and data analysis for separation and classification. Sparsity constraints are commonly set on the model to promote discovery of a small number of dominant patterns.
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::e5a47b8620af4d844ca55a29a6db8c2a
https://doi.org/10.1109/icassp.2015.7178807
https://doi.org/10.1109/icassp.2015.7178807
Autor:
Antti Hurmalainen, Tuomas Virtanen
Publikováno v:
ASRU
Non-negative spectral factorisation with long temporal context has been successfully used for noise robust recognition of speech in multi-source environments. Sparse classification from activations of speech atoms can be employed instead of conventio
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::6fa8ddaf1a78013a5f0b5bf17f013517
https://doi.org/10.1109/asru.2013.6707724
https://doi.org/10.1109/asru.2013.6707724
Modelling spectro-temporal dynamics in factorisation-based noise-robust automatic speech recognition
Autor:
Tuomas Virtanen, Antti Hurmalainen
Publikováno v:
ICASSP
Non-negative spectral factorisation has been used successfully for separation of speech and noise in automatic speech recognition, both in feature-enhancing front-ends and in direct classification. In this work, we propose employing spectro-temporal
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::f45c8e7e3439cb6b4cb38fb106d9b9fb
https://doi.org/10.1109/icassp.2012.6288823
https://doi.org/10.1109/icassp.2012.6288823
Publikováno v:
ICASSP
High noise robustness has been achieved in speech recognition by using sparse exemplar-based methods with spectrogram windows spanning up to 300 ms. A downside is that a large exemplar dictionary is required to cover sufficiently many spectral patter
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::c7bcda7fb7cd36d01b8a119cbee37a99
https://doi.org/10.1109/icassp.2011.5947376
https://doi.org/10.1109/icassp.2011.5947376
Publikováno v:
IEEE Transactions on Audio, Speech, and Language Processing
This paper proposes to use exemplar-based sparse representations for noise robust automatic speech recognition. First, we describe how speech can be modeled as a linear combination of a small number of exemplars from a large speech exemplar dictionar
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::20fef9de0f7a8b57a2572f5c6e7d6f9c
http://hdl.handle.net/2066/94410
http://hdl.handle.net/2066/94410