Combining multi-band and frequency-filtering techniques for speech recognition in noisy environments

While current speech recognisers give acceptable performance in carefully controlled environments, their performance degrades rapidly when they are applied in more realistic situations. Generally, the environmental noise may be classified into two classes: the wide-band noise and narrow band noise. While the multi-band model has been shown to be capable of dealing with speech corrupted by narrow-band noise, it is ineffective for wide-band noise. In this paper, we suggest a combination of the frequency-filtering technique with the probabilistic union model in the multi-band approach. The new system has been tested on the TIDIGITS database, corrupted by white noise, noise collected from a railway station, and narrow-band noise, respectively. The results have shown that this approach is capable of dealing with noise of narrow-band or wide-band characteristics, assuming no knowledge about the noisy environment.

Robust bimodal person identification using face and speech with limited training data and corruption of both modalities

This paper presents a novel method of audio-visual fusion for person identification where both the speech and facial modalities may be corrupted, and there is a lack of prior knowledge about the corruption. Furthermore, we assume there is a limited amount of training data for each modality (e.g., a short training speech segment and a single training facial image for each person). A new representation and a modified cosine similarity are introduced for combining and comparing bimodal features with limited training data as well as vastly differing data rates and feature sizes. Optimal feature selection and multicondition training are used to reduce the mismatch between training and testing, thereby making the system robust to unknown bimodal corruption. Experiments have been carried out on a bimodal data set created from the SPIDRE and AR databases with variable noise corruption of speech and occlusion in the face images. The new method has demonstrated improved recognition accuracy.

An iterative longest matching segment approach to speech enhancement with additive noise and channel distortion

This paper presents a new approach to speech enhancement from single-channel measurements involving both noise and channel distortion (i.e., convolutional noise), and demonstrates its applications for robust speech recognition and for improving noisy speech quality. The approach is based on finding longest matching segments (LMS) from a corpus of clean, wideband speech. The approach adds three novel developments to our previous LMS research. First, we address the problem of channel distortion as well as additive noise. Second, we present an improved method for modeling noise for speech estimation. Third, we present an iterative algorithm which updates the noise and channel estimates of the corpus data model. In experiments using speech recognition as a test with the Aurora 4 database, the use of our enhancement approach as a preprocessor for feature extraction significantly improved the performance of a baseline recognition system. In another comparison against conventional enhancement algorithms, both the PESQ and the segmental SNR ratings of the LMS algorithm were superior to the other methods for noisy speech enhancement.

Speech Enhancement from Additive Noise and Channel Distortion - a Corpus-Based Approach

This paper presents a new approach to single-channel speech enhancement involving both noise and channel distortion (i.e., convolutional noise). The approach is based on finding longest matching segments (LMS) from a corpus of clean, wideband speech. The approach adds three novel developments to our previous LMS research. First, we address the problem of channel distortion as well as additive noise. Second, we present an improved method for modeling noise. Third, we present an iterative algorithm for improved speech estimates. In experiments using speech recognition as a test with the Aurora 4 database, the use of our enhancement approach as a preprocessor for feature extraction significantly improved the performance of a baseline recognition system. In another comparison against conventional enhancement algorithms, both the PESQ and the segmental SNR ratings of the LMS algorithm were superior to the other methods for noisy speech enhancement. Index Terms: corpus-based speech model, longest matching segment, speech enhancement, speech recognition

Merit, Speech Disability and Discrimination

There is a substantial body of evidence – going back over decades – which indicates that the employment sphere is difficult for those who suffer a speech disability. To a large extent, I argue, this is due to the setting of merit in terms of orality and aesthetic. It also relates to the low perception of competence of the speech disabled. I argue that to be effective against discrimination the notion of merit and its assessment requires focus. ‘Merit’ as a concept in discrimination law has had its critics, yet it remains important to investigate it as social construct in order to help understand discrimination and how to counter this. For example, in this article I look at an instance where the resetting of what was viewed as ‘meritorious’ in judicial recruitment successfully improved the diversity in lower judicial posts.

Further, given the relative failure of the employment tribunal system to improve the general position of those who are disabled, I look to alternative methods of countering disability discrimination. The suggestion provided is that an enforced ombudsman type approach capable of dealing with what may be the core issue around employment discrimination (‘merit’) would provide a better mechanism for handling the general situation of disability discrimination than the tribunal system.

Wide Matching - An Approach to Improving Noise Robustness for Speech Enhancement

It is shown that under certain conditions it is possible to obtain a good speech estimate from noise without requiring noise estimation. We study an implementation of the theory, namely wide matching, for speech enhancement. The new approach performs sentence-wide joint speech segment estimation subject to maximum recognizability to gain noise robustness. Experiments have been conducted to evaluate the new approach with variable noises and SNRs from -5 dB to noise free. It is shown that the new approach, without any estimation of the noise, significantly outperformed conventional methods in the low SNR conditions while retaining comparable performance in the high SNR conditions. It is further suggested that the wide matching and deep learning approaches can be combined towards a highly robust and accurate speech estimator.

Brain responses and looking behavior during audiovisual speech integration in infants predict auditory speech comprehension in the second year of life

The use of visual cues during the processing of audiovisual (AV) speech is known to be less efficient in children and adults with language difficulties and difficulties are known to be more prevalent in children from low-income populations. In the present study, we followed an economically diverse group of thirty-seven infants longitudinally from 6–9 months to 14–16 months of age. We used eye-tracking to examine whether individual differences in visual attention during AV processing of speech in 6–9 month old infants, particularly when processing congruent and incongruent auditory and visual speech cues, might be indicative of their later language development. Twenty-two of these 6–9 month old infants also participated in an event-related potential (ERP) AV task within the same experimental session. Language development was then followed-up at the age of 14–16 months, using two measures of language development, the Preschool Language Scale and the Oxford Communicative Development Inventory. The results show that those infants who were less efficient in auditory speech processing at the age of 6–9 months had lower receptive language scores at 14–16 months. A correlational analysis revealed that the pattern of face scanning and ERP responses to audiovisually incongruent stimuli at 6–9 months were both significantly associated with language development at 14–16 months. These findings add to the understanding of individual differences in neural signatures of AV processing and associated looking behavior in infants.

Brain responses to audiovisual speech mismatch in infants are associated with individual differences in looking behaviour

Research on audiovisual speech integration has reported high levels of individual variability, especially among young infants. In the present study we tested the hypothesis that this variability results from individual differences in the maturation of audiovisual speech processing during infancy. A developmental shift in selective attention to audiovisual speech has been demonstrated between 6 and 9 months with an increase in the time spent looking to articulating mouths as compared to eyes (Lewkowicz & Hansen-Tift. (2012) Proc. Natl Acad. Sci. USA, 109, 1431–1436; Tomalski et al. (2012) Eur. J. Dev. Psychol., 1–14). In the present study we tested whether these changes in behavioural maturational level are associated with differences in brain responses to audiovisual speech across this age range. We measured high-density event-related potentials (ERPs) in response to videos of audiovisually matching and mismatched syllables /ba/ and /ga/, and subsequently examined visual scanning of the same stimuli with eye-tracking. There were no clear age-specific changes in ERPs, but the amplitude of audiovisual mismatch response (AVMMR) to the combination of visual /ba/ and auditory /ga/ was strongly negatively associated with looking time to the mouth in the same condition. These results have significant implications for our understanding of individual differences in neural signatures of audiovisual speech processing in infants, suggesting that they are not strictly related to chronological age but instead associated with the maturation of looking behaviour, and develop at individual rates in the second half of the first year of life.

Speech of Mr. C. G. Memminger, on the question of rechartering the bank of the state of South Carolina: delivered in the House of Representatives, December 9, 1848

This speech by Mr. Memminger offers resolutions on the issue of rechartering the bank of the state of South Carolina. The issues presented are the Bank of the State is founded on an erroneous policy, unwise for a state to engage in banking, not practical to recharter the Bank of the State and a special committee of each house should be appointed to advise how to carry out these resolutions at the next session.

The sectionalism of the tariff bill. Speech of Hon. John L. McLaurin, of South Carolina, in the Senate of the United States

This document contains a speech by John L. McLaurin, representative of South Carolina. Sections of the speech include: sectionalism exposed, the bill might have been defeated, the south plundered of its rights, not a protectionist, fraudulent demands of New England, Hon. Randolph Tucker, Hon. W.R. Morrison, and Hon. R.Q. Mills strangers to the doctrine in 1882, a tariff for revenue against the doctrine of free raw material, don’t want Cleveland’s interpretation, contest of schedules, and my remedy.

Speech of Hon. John Lowndes McLaurin, of South Carolina, in the Senate of the United States, Monday, December 9, 1901

This document contains a speech by John L. McLaurin of South Carolina presented in the Senate of the United States. Sections of the speech include: sectionalism the cause, conditions in South Carolina, the federal administration in South Carolina, should not array class against class, freedom of thought and speech, the issues, under caucus dictation the Senate no longer a deliberative body, the beginning of the fight, matter of arraying class against class, freedom of thought and speech, and issues.

The tariff. Speech of Representative J. L. McLaurin, of South Carolina, in the House of Representatives, Tuesday, March 23, 1897

This document contains a speech of John L. McLaurin, representative of South Carolina, in the House of Representatives on Tuesday, March 23, 1897 about proposed tariffs.

Where the consumer is not a myth : letters from England describing economic and industrial conditions by Zach McGhee, Washington correspondent : speech of Hon. Wyatt Aiken in the House of Representatives, Tuesday, March 22, 1910

This document contains a speech of David Wyatt Aiken, representative of South Carolina, to the House of Representatives on Tuesday, March 22, 1910. Much of the speech is a letter from Zach McGhee, Washington correspondent of The State newspaper on industrial conditions in England and Europe.

Protection to farmers : speech of Hon. D. Wyatt Aiken, of South Carolina, in the House of Representatives, Tuesday, March 6, 1882, upon the Agricultural Appropriation Bill

This speech is about the Agricultural Appropriation Bill, which is a bill making appropriations for the Agricultural Department of the Government for the fiscal year ending in June 30,1883. Mr. Aiken is approved by the chairman to speak and he goes on to explain that he agrees with the majority of the bill with the exception of two or three clauses. He gives the reasoning behind his objections in the rest of the speech.