The use of phase in complex spectrum subtraction for robust speech recognition

In this paper we propose a new method for utilising phase information by complementing it with traditional magnitude-only spectral subtraction speech enhancement through Complex Spectrum Subtraction (CSS). The proposed approach has the following advantages over traditional magnitude-only spectral subtraction: (a) it introduces complementary information to the enhancement algorithm; (b) it reduces the total number of algorithmic parameters, and; (c) is designed for improving clean speech magnitude spectra and is therefore suitable for both automatic speech recognition (ASR) and speech perception applications. Oracle-based ASR experiments verify this approach, showing an average of 20% relative word accuracy improvements when accurate estimates of the phase spectrum are available. Based on sinusoidal analysis and assuming stationarity between observations (which is shown to be better approximated as the frame rate is increased), this paper also proposes a novel method for acquiring the phase information called Phase Estimation via Delay Projection (PEDEP). Further oracle ASR experiments validate the potential for the proposed PEDEP technique in ideal conditions. Realistic implementation of CSS with PEDEP shows performance comparable to state of the art spectral subtraction techniques in a range of 15-20 dB signal-to-noise ratio environments. These results clearly demonstrate the potential for using phase spectra in spectral subtractive enhancement applications, and at the same time highlight the need for deriving more accurate phase estimates in a wider range of noise conditions.

Microphone array beamforming approach to blind speech separation

In this paper, we present a microphone array beamforming approach to blind speech separation. Unlike previous beamforming approaches, our system does not require a-priori knowledge of the microphone placement and speaker location, making the system directly comparable other blind source separation methods which require no prior knowledge of recording conditions. Microphone location is automatically estimated using an assumed noise field model, and speaker locations are estimated using cross correlation based methods. The system is evaluated on the data provided for the PASCAL Speech Separation Challenge 2 (SSC2), achieving a word error rate of 58% on the evaluation set.

Low-cost hardware speech enhancement for improved speech recognition in automotive environments

Voice recognition is one of the key enablers to reduce driver distraction as in-vehicle systems become more and more complex. With the integration of voice recognition in vehicles, safety and usability are improved as the driver’s eyes and hands are not required to operate system controls. Whilst speaker independent voice recognition is well developed, performance in high noise environments (e.g. vehicles) is still limited. La Trobe University and Queensland University of Technology have developed a low-cost hardware-based speech enhancement system for automotive environments based on spectral subtraction and delay–sum beamforming techniques. The enhancement algorithms have been optimised using authentic Australian English collected under typical driving conditions. Performance tests conducted using speech data collected under variety of vehicle noise conditions demonstrate a word recognition rate improvement in the order of 10% or more under the noisiest conditions. Currently developed to a proof of concept stage there is potential for even greater performance improvement.

Recognising audio-visual speech in vehicles using the AVICAR database

Interacting with technology within a vehicle environment using a voice interface can greatly reduce the effects of driver distraction. Most current approaches to this problem only utilise the audio signal, making them susceptible to acoustic noise. An obvious approach to circumvent this is to use the visual modality in addition. However, capturing, storing and distributing audio-visual data in a vehicle environment is very costly and difficult. One current dataset available for such research is the AVICAR [1] database. Unfortunately this database is largely unusable due to timing mismatch between the two streams and in addition, no protocol is available. We have overcome this problem by re-synchronising the streams on the phone-number portion of the dataset and established a protocol for further research. This paper presents the first audio-visual results on this dataset for speaker-independent speech recognition. We hope this will serve as a catalyst for future research in this area.

Maternal administration of erythromycin fails to eradicate intrauterine ureaplasma infection in an ovine model

Erythromycin is the standard antibiotic used for treatment of Ureaplasma species during 3 pregnancy; however, maternally administered erythromycin may be ineffective at eliminating 4 intra-amniotic ureaplasma infections. We asked if erythromycin would eradicate intra-amniotic 5 ureaplasma infections in pregnant sheep. At 50 days of gestation (d, term=150d) pregnant ewes 6 received intra-amniotic injections of erythromycin-sensitive U. parvum serovar 3 (n=16) or 10B 7 medium (n=16). At 100d, amniocentesis was performed; five fetal losses (ureaplasma group: 8 n=4; 10B group: n=1) had occurred by this time. Remaining ewes were allocated into treatment 9 subgroups: medium only (M, n=7); medium and erythromycin (M/E, n=8); ureaplasma only (Up, 10 n=6) or ureaplasma and erythromycin (Up/E, n=6). Erythromycin was administered intra11 muscularly (500 mg), eight-hourly for four days (100d-104d). Amniotic fluid samples were 12 collected at 105d. At 125d preterm fetuses were surgically delivered and specimens were 13 collected for culture and histology. Erythromycin was quantified in amniotic fluid by liquid 14 chromatography-mass spectrometry. Ureaplasmas were isolated from the amniotic fluid, 15 chorioamnion and fetal lung of animals from the Up and Up/E groups, however, the numbers of 16 U. parvum recovered were not different between these groups. Inflammation in the 17 chorioamnion, cord and fetal lung was increased in ureaplasma-exposed animals compared to 18 controls, but was not different between the Up and Up/E groups. Erythromycin was detected in 19 amniotic fluid samples, although concentrations were low (<10-76 ng/mL). This study 20 demonstrates that maternally administered erythromycin does not eradicate chronic, intra- amniotic ureaplasma infections or improve fetal outcomes in an ovine model, potentially due to 22 the poor placental passage of erythromycin.

Observational assessment and maternal reports of motivation in children and adolescents with Down Syndrome

Despite a lack of consistent empirical evidence, there has been an ongoing assumption that intellectual disability is associated with reduced levels of motivation. The participants in this study were 33 children with Down syndrome ages 10–15 years and 33 typically developing 3–8-year-old children. Motivation was measured through observational assessments of curiosity, preference for challenge, and persistence, as well as maternal reports. There were no significant group differences on motivation tasks, but mothers of children with Down syndrome rated their children significantly lower on motivation than did parents of typically developing children. There were some intriguing group differences in the pattern of correlations among observations and parent reports. The findings challenge long-held views that individuals with intellectual disability are invariably deficient in motivation.

Maternal exposure to ambient temperature and the risk of preterm birth and stillbirth in Brisbane, Australia

Almost 10% of all births are preterm and 2.2% are stillbirths globally. Recent research has suggested that environmental factors may be a contributory cause to these adverse birth outcomes. The authors examined the relationship between ambient temperature and preterm birth and stillbirth in Brisbane, Australia between 2005 and 2009 (n = 101,870). They used a Cox proportional hazard model with live birth and stillbirth as competing risks. They also examined if there were periods of the pregnancy where exposure to high temperatures had a greater effect. Exposure to higher ambient temperatures during pregnancy increased the risk of stillbirth. The hazard ratio for stillbirth was 0.3 at 12 °C relative to the reference temperature at 21 °C. The temperature effect was greatest for fetuses of less than 36 weeks of gestation. There was an association between higher temperature and shorter gestation, as the hazard ratio for live birth was 0.96 at 15 °C and 1.02 at 25 °C. This effect was greatest at later gestational ages. The results provide strong evidence of an association between increased temperature and increased risk of stillbirth and shorter gestations.

The influence of maternal ethnic group and diet on breast milk fatty acid composition

INTRODUCTION: Breast milk fatty acids play a major role in infant development. However, no data have compared the breast milk composition of different ethnic groups living in the same environment. We aimed to (i) investigate breast milk fatty acid composition of three ethnic groups in Singapore and (ii) determine dietary fatty acid patterns in these groups and any association with breast milk fatty acid composition. MATERIALS AND METHODS: This was a prospective study conducted at a tertiary hospital in Singapore. Healthy pregnant women with the intention to breastfeed were recruited. Diet profile was studied using a standard validated 3-day food diary. Breast milk was collected from mothers at 1 to 2 weeks and 6 to 8 weeks postnatally. Agilent gas chromatograph (6870N) equipped with a mass spectrometer (5975) and an automatic liquid sampler (ALS) system with a split mode was used for analysis. RESULTS: Seventy-two breast milk samples were obtained from 52 subjects. Analysis showed that breast milk ETA (Eicosatetraenoic acid) and ETA:EA (Eicosatrienoic acid) ratio were significantly different among the races (P = 0.031 and P = 0.020), with ETA being the highest among Indians and the lowest among Malays. Docosahexaenoic acid was significantly higher among Chinese compared to Indians and Malays. No difference was demonstrated in n3 and n6 levels in the food diet analysis among the 3 ethnic groups. CONCLUSIONS: Differences exist in breast milk fatty acid composition in different ethnic groups in the same region, although no difference was demonstrated in the diet analysis. Factors other than maternal diet may play a role in breast milk fatty acid composition.

Improving visual noise insensitivity in small vocabulary audio visual speech recognition applications

Visual noise insensitivity is important to audio visual speech recognition (AVSR). Visual noise can take on a number of forms such as varying frame rate, occlusion, lighting or speaker variabilities. The use of a high dimensional secondary classifier on the word likelihood scores from both the audio and video modalities is investigated for the purposes of adaptive fusion. Preliminary results are presented demonstrating performance above the catastrophic fusion boundary for our confidence measure irrespective of the type of visual noise presented to it. Our experiments were restricted to small vocabulary applications.

Speech recognition in adverse environments using lip information

The performance of automatic speech recognition systems deteriorates in the presence of noise. One known solution is to incorporate video information with an existing acoustic speech recognition system. We investigate the performance of the individual acoustic and visual sub-systems and then examine different ways in which the integration of the two systems may be performed. The system is to be implemented in real time on a Texas Instruments' TMS320C80 DSP.

Robust speaker verification via fusion of speech and lip modalities

This paper investigates the use of lip information, in conjunction with speech information, for robust speaker verification in the presence of background noise. It has been previously shown in our own work, and in the work of others, that features extracted from a speaker's moving lips hold speaker dependencies which are complementary with speech features. We demonstrate that the fusion of lip and speech information allows for a highly robust speaker verification system which outperforms the performance of either sub-system. We present a new technique for determining the weighting to be applied to each modality so as to optimize the performance of the fused system. Given a correct weighting, lip information is shown to be highly effective for reducing the false acceptance and false rejection error rates in the presence of background noise

The use of temporal speech and lip information for multi-modal speaker identification via multi-stream HMMs

Investigates the use of temporal lip information, in conjunction with speech information, for robust, text-dependent speaker identification. We propose that significant speaker-dependent information can be obtained from moving lips, enabling speaker recognition systems to be highly robust in the presence of noise. The fusion structure for the audio and visual information is based around the use of multi-stream hidden Markov models (MSHMM), with audio and visual features forming two independent data streams. Recent work with multi-modal MSHMMs has been performed successfully for the task of speech recognition. The use of temporal lip information for speaker identification has been performed previously (T.J. Wark et al., 1998), however this has been restricted to output fusion via single-stream HMMs. We present an extension to this previous work, and show that a MSHMM is a valid structure for multi-modal speaker identification

The use of speech and lip modalities for robust speaker verification under adverse conditions

Investigates the use of lip information, in conjunction with speech information, for robust speaker verification in the presence of background noise. We have previously shown (Int. Conf. on Acoustics, Speech and Signal Proc., vol. 6, pp. 3693-3696, May 1998) that features extracted from a speaker's moving lips hold speaker dependencies which are complementary with speech features. We demonstrate that the fusion of lip and speech information allows for a highly robust speaker verification system which outperforms either subsystem individually. We present a new technique for determining the weighting to be applied to each modality so as to optimize the performance of the fused system. Given a correct weighting, lip information is shown to be highly effective for reducing the false acceptance and false rejection error rates in the presence of background noise

Can audio-visual speech recognition outperform acoustically enhanced speech recognition in automotive environment?

The use of visual features in the form of lip movements to improve the performance of acoustic speech recognition has been shown to work well, particularly in noisy acoustic conditions. However, whether this technique can outperform speech recognition incorporating well-known acoustic enhancement techniques, such as spectral subtraction, or multi-channel beamforming is not known. This is an important question to be answered especially in an automotive environment, for the design of an efficient human-vehicle computer interface. We perform a variety of speech recognition experiments on a challenging automotive speech dataset and results show that synchronous HMM-based audio-visual fusion can outperform traditional single as well as multi-channel acoustic speech enhancement techniques. We also show that further improvement in recognition performance can be obtained by fusing speech-enhanced audio with the visual modality, demonstrating the complementary nature of the two robust speech recognition approaches.

Breastfeeding self-efficacy and alternative techniques to overcome maternal or infant breastfeeding challenges : a retrospective descriptive study

Background: Breastfeeding is the internationally accepted ideal in infant feeding. Ensuring mothers and babies receive optimal benefits, in both the short and long term, is dependent upon the successful establishment of breastfeeding in the first week. Many maternal and infant challenges can occur during the establishment of breastfeeding (Lactogenesis II). There are also many methods and devices (alternative techniques) which can be used to help, but the majority do not have an evidence-base. The mother.s self-confidence (self-efficacy) can be challenged by these unexpected circumstances, but understanding of the relationship is unclear. Method: This descriptive study used mail survey (including the Breastfeeding Self-Efficacy Scale . Short Form) to obtain the mother.s reports of their self-efficacy and their breastfeeding experience during the first week following birth, as well as actual use of alternative techniques. This study included all mothers of full term healthy singleton infants from one private hospital in Brisbane who began any breastfeeding. The data collection took place from November 2008 to February 2009. Ethical approval was granted from the research site and QUT Human Research Ethics Committee. Results: A total of 128 questionnaires were returned, a response rate of 56.9%. The sample was dissimilar to the Queensland population with regard to age, income, and education level, all of which were higher in this study. The sample was similar to the Queensland population in terms of parity and marital status. The rate of use of alternative techniques was 48.3%. The mean breastfeeding self-efficacy score of those who used any alternative technique was 43.43 (SD=12.19), and for those who did not, it was 58.32 (SD=7.40). Kruskal-Wallis analysis identified that the median self efficacy score for those who used alternative techniques was significantly lower than median self efficacy scores for those who did not use alternative techniques. The reasons women used alternative techniques varied widely, and their knowledge of alternative techniques was good. Conclusion: This study is the first to document breastfeeding self-efficacy of women who used alternative techniques to support their breastfeeding goals in the first week postpartum. An individualised clinical intervention to develop women.s self-efficacy with breastfeeding is important to assist mother/infant dyads encountering challenges to breastfeeding in the first week postpartum.