Publications de Hynek Hermansky

Phase AutoCorrelation (PAC) features for noise robust speech recognition

Mathew Magimai Doss, Shajith Ikbal, Hynek Hermansky, Hemant Misra

In this paper, we introduce a new class of noise robust features derived from an alternative measure of autocorrelation representing the phase variation of speech signal frame over time. These features, referred to as Phase AutoCorrelation (PAC) features i ...

2012

Analysis of MLP Based Hierarchical Phoneme Posterior Probability Estimator

Hervé Bourlard, Mathew Magimai Doss, Hynek Hermansky, Joel Praveen Pinto

We analyze a simple hierarchical architecture consisting of two multilayer perceptron (MLP) classifiers in tandem to estimate the phonetic class conditional probabilities. In this hierarchical setup, the first MLP classifier is trained using standard acous ...

2011

Autoregressive Models of Amplitude Modulations in Audio Compression

Hynek Hermansky, Sriram Ganapathy, Petr Motlicek

We present a scalable medium bit-rate wide-band audio coding technique based on frequency domain linear prediction (FDLP). FDLP is an efficient method for representing the long-term amplitude modulations of speech/audio signals using autoregressive models. ...

2010

Wide-Band Audio Coding based on Frequency Domain Linear Prediction

Hynek Hermansky, Sriram Ganapathy, Petr Motlicek

We revisit an original concept of speech coding in which the signal is separated into the carrier modulated by the signal envelope. A recently developed technique, called frequency-domain linear prediction (FDLP), is applied for the efficient estimation of ...

2010

Recognition Of Reverberant Speech Using Frequency Domain Linear Prediction

Hynek Hermansky, Sriram Ganapathy, Samuel Thomas

Performance of a typical automatic speech recognition (ASR) system severely degrades when it encounters speech from reverberant environments. Part of the reason for this degradation is the feature extraction techniques that use analysis windows which are m ...

IEEE Institute of Electrical and Electronics Engineers2008

Modulation Frequency Features For Phoneme Recognition In Noisy Speech

Hynek Hermansky, Sriram Ganapathy, Samuel Thomas

In this letter, a new feature extraction technique based on modulation spectrum derived from syllable-length segments of sub-band temporal envelopes is proposed. These sub-band envelopes are derived from auto-regressive modelling of Hilbert envelopes of th ...

2008

Using Comparison of Parallel Phoneme Probability streams for OOV Word Detection

Mathew Magimai Doss, Hynek Hermansky, Tamara Tosic

In this paper, we investigate the approach of comparing two different parallel streams of phoneme posterior probability estimates for OOV word detection. The ﬁrst phoneme posterior probability stream is estimated using only the knowledge of short-term acou ...

2008

Emulating Temporal Receptive Fields of Higher Level Auditory Neurons for ASR

Hynek Hermansky

This paper proposes modifications to the Multi-resolution RASTA (MRASTA) feature extraction technique for the automatic speech recognition (ASR). By emulating asymmetries of the temporal receptive field (TRF) profiles of higher level auditory neurons, we o ...

Springer-Verlag New York, Ms Ingrid Cunningham, 175 Fifth Ave, New York, Ny 10010 Usa2008