In this paper, we present our efforts towards developing a robust automatic speaker verification (ASV) system for children when the domain-specific data is limited. For that purpose, we have studied the effect of in-domain and out-of-domain data augmentation. Several different combinations of data augmentation are studied in this work. Speed and pitch perturbation of children's speech are employed for synthetically creating in-domain data to be used for augmentation. For out-of-domain data augmentation, on the other hand, adults' speech is pooled together with children's speech. At the same time, voice conversion (VC) is also applied on adults' speech to alter the acoustic attributes. VC of adults' speech makes it perceptually similar to that of children's speech. The converted adults' data is then used for augmentation. The ASV systems developed in this study employ x-vectors derived using a time-delay deep neural network. In addition to that, probabilistic linear discriminant analysis is used for scoring the performance. The explored methods of data augmentation are noted to reduce the equal error rate as well as minimum decision cost function by a large margin.

We address the problem of estimating the time-varying spectral envelope of a speech signal using a spectro-temporal demodulation technique. Unlike the conventional spectrogram, we consider a pitch-adaptive spectrogram and model a spectro-temporal patch using an amplitude- and frequency-modulated two-dimensional (2-D) cosine signal. We employ a demodulation technique based on the Riesz transform that we proposed recently to estimate the amplitude and frequency modulations. The amplitude modulation (AM) corresponds to the vocal-tract filter magnitude response (or envelope) and the frequency modulation (FM) corresponds to the excitation. We consider the AM and demonstrate its effectiveness by incorporating it as an acoustic feature for local conditioning in the statistical WaveNet vocoder for the task of speech synthesis. The quality of the synthesized speech obtained with the Riesz envelope is compared with that obtained using the envelope estimated by the WORLD vocoder. Objective measures and subjective listening tests on the CMU-Arctic database show that the quality of synthesis is superior to that obtained using the WORLD envelope. This study thus establishes the Riesz envelope as an efficient alternative to the WORLD envelope.

In this paper, we suggest a new parallel, non-causal and shallow waveform domain architecture for speech enhancement based on FFTNet, a neural network for generating high quality audio waveform. In contrast to other waveform based approaches like WaveNet, FFTNet uses an initial wide dilation pattern. Such an architecture better represents the long term correlated structure of speech in the time domain, where noise is usually highly non-correlated, and therefore it is suitable for waveform domain based speech enhancement. To further strengthen this feature of FFTNet, we suggest a non-causal FFTNet architecture, where the present sample in each layer is estimated from the past and future samples of the previous layer. By suggesting a shallow network and applying non-causality within certain limits, the suggested FFTNet for speech enhancement (SE-FFTNet) uses much fewer parameters compared to other neural network based approaches for speech enhancement like WaveNet and SEGAN. Specifically, the suggested network has considerably reduced model parameters: 32% fewer compared to WaveNet and 87% fewer compared to SEGAN. Finally, based on subjective and objective metrics, SE-FFTNet outperforms WaveNet in terms of enhanced signal quality, while it provides equally good performance as SEGAN. A Tensorflow implementation of the architecture is provided at 1 .

The presence of velopharyngeal dysfunction in individuals with cleft palate (CP) nasalizes the voiced stops. Due to this, voiced stops (/b/, /d/, /g/) tend to be perceive like nasal consonants (/m/, /n/, /ng/). In this work, a novel algorithm is proposed for the detection of nasalized voiced stops in CP speech using epoch-synchronous features. Speech regions corresponding to consonant and consonant-vowel transitions are segmented using the knowledge of glottal activity, syllable nucleus, low-frequency spectral dominance, and vowel onset point. The segmented regions are epoch-synchronously processed to analyze the spectral, spectro-temporal, excitation source, and periodicity characteristics of normal and nasalized voiced stops. Spectral and spectro temporal features are computed using single pole filter based time-frequency representation. The amplitude of Hilbert envelope of linear prediction residual, measured around the epoch is used to analyze the effect of nasalization on excitation source. Comparison of speech frames of successive inter-epoch intervals is carried out to analyze the periodicity characteristics. The proposed features are used to develop a support vector machine classifier for the classification of normal and nasalized voiced stops. Segmentation accuracy for the proposed knowledge based method is found to be better than the hidden Markov model based force-alignment approach. The detection rate of nasalized voiced stops is found to be high for the proposed epoch synchronous features than the conventional Mel-frequency cepstral coefficients.

In this paper, we explore the possibility of using the WaveNet architecture as a statistical vocoder. In that case, the generation of speech waveforms is locally conditioned only by acoustic features. Focusing on the single speaker case at the moment, we investigate the impact of the local conditions as well as that of the amount of data available for training. Furthermore, variations of the WaveNet architecture are considered and discussed in the context of our work. We compare our work against a very recent work which also used WaveNet architecture as a speech vocoder using the same speech data. More specifically, we used two female and two male speakers from the CMU-ARCTIC database to contrast the use of cepstrum coefficients and filter-bank features as local conditioners with the goal to improve the overall quality for both male and female speakers. In the paper we also discuss the impact of the size of the training data. Objective metrics for quality and intelligibility of the generated by the WaveNet speech as well as subjective tests support our suggestions.

In this paper, we have explored the role of combining prosodic variables with the existing acoustic features in the context of children's speech recognition using acoustic models trained on adults' speech. The explored acoustic features are Mel-frequency cepstral coefficients (MFCC) and perceptual linear prediction cepstral coefficients (PLPCC) while the considered prosodic variables are loudness, voice-intensity and voice-probability. An analysis presented in this paper shows that, given that the textual content remains the same, the considered prosodic variables exhibit very similar contours for adults' and children's speech. At the same time, the contours differ a lot when the context is different. Consequently, inclusion of prosodic information reduces the inter-speaker differences and increases the class discrimination. This subsequently improves the recognition performance. Further improvements are obtained by projecting the feature vectors obtained by combining the two features to a lower-dimensional subspace. The same has been experimentally verified in this study for mismatched speech recognition using deep neural network (DNN) based system. On combining MFCC (PLPCC) and prosodic features, a relative improvement of 16% (14%) is noted on decoding children's speech using adult data trained DNN models.

We consider a two-dimensional demodulation framework for spectro-temporal analysis of the speech signal. We construct narrowband (NB) speech spectrograms, and demodulate them using the Riesz transform, which is a two-dimensional extension of the Hilbert transform. The demodulation results in timefrequency envelope (amplitude modulation or AM) and timefrequency carrier (frequency modulation or FM). The AM corresponds to the vocal tract and is referred to as the vocal tract spectrogram. The FM corresponds to the underlying excitation and is referred to as the carrier spectrogram. The carrier spectrogram exhibits a high degree of time-frequency consistency for voiced sounds. For unvoiced sounds, such a structure is lacking. In addition, the carrier spectrogram reflects the fundamental frequency (F0) variation of the speech signal. We develop a technique to determine the F0 from the carrier spectrogram. The time-frequency consistency is used to determine which time-frequency regions correspond to voiced segments. Comparisons with the state-of-the-art F0 estimation algorithms show that the proposed F0 estimator has high accuracy for telephone channel speech and is robust to noise.

Transcribing children's speech using acoustic models trained on adults' speech is very challenging. In such conditions, a highly degraded recognition performance is reported due to large mismatch in the acoustic/linguistic attributes of the training and test data. The differences in pitch (or fundamental frequency) between the two groups of speakers is one among several mismatch factors. Another important mismatch factor is the difference in speaking rates. To overcome these two sources of mismatch, prosody modification is explored in this letter. Prosody modification is done by using glottal closure instants (GCIs) as anchoring points. The GCIs, in turn, are determined using zero-frequency filtering (ZFF). The ZFF-GCI-based prosody modification is fast and results in highly accurate scaling of pitch and speaking rate. The experimental evaluations studying the effect of prosody modification resulted in a relative improvement of 50% over the baseline.

In this work, new method of source modeling for HMM based speech synthesis is proposed using integrated LP residual (ILPR). The nature of ILPR waveform resembles the glottal flow derivative signal and may keep the speaker characteristics in a better way. The ILPR signal is modeled in the frequency domain by dividing the spectrum into two bands to characterize harmonic and noise components of the voice speech segment. The harmonic components of ILPR signals below the maximum voiced frequency (fm) is modeled using mel-cepstral coefficients called as RMCEPs, whereas noise component above fmis modeled by pitch adaptive triangular noise envelope weighted by the strength of excitation (SoE). The RMCEPs and SoE are modeled on the HMM framework along with MCEPs and F0representing vocal tract information and fundamental frequency, respectively. The synthesized speech by the proposed source modeling reduces the buzziness and improves the speaker similarity compared to the conventional impulse / noise and mixed excitation source modeling and comparable with STRAIGHT based excitation. This is further reflected in both objective and subjective valuations.

The major activity during speech production is glottal activity and is earlier detected using strength of excitation (SoE). This work uses the normalized autocorrelation peak strength (NAPS) and higher order statistics (HOS) as additional features for detecting glottal activity. The three features, namely, SoE, NAPS, and HOS, are, respectively indicators of different attributes of glottal activity, namely, energy, periodicity, and asymmetrical nature of the resulting source signal. The effectiveness of these features is analyzed using the differential electroglottograph signal, zero-frequency filtered signal, and integrated linear prediction residual, as representatives of source signal. The combination of glottal activity information from the three features outperforms any single of them, demonstrating different information represented by each of these features.