Houman Ghaemmaghami

Normal probability testing of snore signals for diagnosis of obstructive sleep apnea

Obstructive Sleep Apnea (OSA) is a highly prevalent disease in which upper airways are collapsed during sleep, leading to serious consequences. The standard method of OSA diagnosis is known as Polysomnography (PSG), which requires an overnight stay in a specifically equipped facility, connected to over 15 channels of measurements. PSG requires (i) contact instrumentation and, (ii) the expert human scoring of a vast amount of data based on subjective criteria. PSG is expensive, time consuming and is difficult to use in community screening or pediatric assessment. Snoring is the most common symptom of OSA. Despite the vast potential, however, it is not currently used in the clinical diagnosis of OSA. In this paper, we propose a novel method of snore signal analysis for the diagnosis of OSA. The method is based on a novel feature that quantifies the non-Gaussianity of individual episodes of snoring. The proposed method was evaluated using overnight clinical snore sound recordings of 86 subjects. The recordings were made concurrently with routine PSG, which was used to establish the ground truth via standard clinical diagnostic procedures. The results indicated that the developed method has a detectability accuracy of 97.34%.

Noise robust voice activity detection using normal probability testing and time-domain histogram analysis

This paper presents a method of voice activity detection (VAD) suitable for high noise scenarios, based on the fusion of two complementary systems. The first system uses a proposed non-Gaussianity score (NGS) feature based on normal probability testing. The second system employs a histogram distance score (HDS) feature that detects changes in the signal through conducting a template-based similarity measure between adjacent frames. The decision outputs by the two systems are then merged using an open-by-reconstruction fusion stage. Accuracy of the proposed method was compared to several baseline VAD methods on a database created using real recordings of a variety of high-noise environments.

Speech Endpoint Detection Using Gradient Based Edge Detection Techniques

This paper proposes a novel method for speech endpoint detection. The developed method utilises gradient based edge detection algorithms, used in image processing, to detect boundaries of continuous speech in noisy conditions. It is simple and has low computational complexity. The accuracy of the proposed method was evaluated and compared to the ITU-T G.729 Annex-B voice activity detection (VAD) algorithm. To do this, the two algorithms were tested using a synthetically produced noisy-speech database, consisting of noisy-speech signals at various lengths and SNR. The results indicated that the developed method outperforms the G.729-B VAD algorithm at various signal-to-noise ratios.

Speakers In The Wild (SITW): The QUT Speaker Recognition System.

This paper presents the QUT speaker recognition system, as a competing system in the Speakers In The Wild (SITW) speaker recognition challenge. Our proposed system achieved an overall ranking of second place, in the main core-core condition evaluations of the SITW challenge. This system uses an ivector/ PLDA approach, with domain adaptation and a deep neural network (DNN) trained to provide feature statistics. The statistics are accumulated by using class posteriors from the DNN, in place of GMM component posteriors in a typical GMM UBM i-vector/PLDA system. Once the statistics have been collected, the i-vector computation is carried out as in a GMM-UBM based system. We apply domain adaptation to the extracted i-vectors to ensure robustness against dataset variability, PLDA modelling is used to capture speaker and session variability in the i-vector space, and the processed i-vectors are compared using the batch likelihood ratio. The final scores are calibrated to obtain the calibrated likelihood scores, which are then used to carry out speaker recognition and evaluate the performance of the system. Finally, we explore the practical application of our system to the core-multi condition recordings of the SITW data and propose a technique for speaker recognition in recordings with multiple speakers.

Acoustic Adaptation in Cross Database Audio Visual SHMM Training for Phonetic Spoken Term Detection

Visual information in the form of lip movements of the speaker has been shown to improve the performance of speech recognition and search applications. In our previous work, we proposed cross database training of synchronous hidden Markov models (SHMMs) to make use of external large and publicly available audio databases in addition to the relatively small given audio visual database. In this work, the cross database training approach is improved by performing an additional audio adaptation step, which enables audio visual SHMMs to benefit from audio observations of the external audio models before adding visual modality to them. The proposed approach outperforms the baseline cross database training approach in clean and noisy environments in terms of phone recognition accuracy as well as spoken term detection (STD) accuracy.