Bootstrapping a statistical speech translator from a rule-based one

We describe a series of experiments in which we start with English to French and English to Japanese versions of an Open Source rule-based speech translation system for a medical domain, and bootstrap correspondign statistical systems. Comparative evaluation reveals that the rule-based systems are still significantly better than the statistical ones, despite the fact that considerable effort has been invested in tuning both the recognition and translation components; also, a hybrid system only marginally improved recall at the cost of a los in precision. The result suggests that rule-based architectures may still be preferable to statistical ones for safety-critical speech translation tasks.

Speaker recognition improvement using blind inversion of distortions

In this paper we propose the inversion of nonlinear distortions in order to improve the recognition rates of a speaker recognizer system. We study the effect of saturations on the test signals, trying to take into account real situations where the training material has been recorded in a controlled situation but the testing signals present some mismatch with the input signal level (saturations). The experimental results shows that a combination of several strategies can improve the recognition rates with saturated test sentences from 80% to 89.39%, while the results with clean speech (without saturation) is 87.76% for one microphone.

Spontaneous Speech and Emotional Response modeling based on One-class classifier oriented to Alzheimer Disease diagnosis

The purpose of our project is to contribute to earlier diagnosis of AD and better estimates of its severity by using automatic analysis performed through new biomarkers extracted from non-invasive intelligent methods. The methods selected in this case are speech biomarkers oriented to Sponta-neous Speech and Emotional Response Analysis. Thus the main goal of the present work is feature search in Spontaneous Speech oriented to pre-clinical evaluation for the definition of test for AD diagnosis by One-class classifier. One-class classifi-cation problem differs from multi-class classifier in one essen-tial aspect. In one-class classification it is assumed that only information of one of the classes, the target class, is available. In this work we explore the problem of imbalanced datasets that is particularly crucial in applications where the goal is to maximize recognition of the minority class as in medical diag-nosis. The use of information about outlier and Fractal Dimen-sion features improves the system performance.

Lexical Influences on Spoken Spondaic Word Recognition in Hearing-Impaired Patients.

Top-down contextual influences play a major part in speech understanding, especially in hearing-impaired patients with deteriorated auditory input. Those influences are most obvious in difficult listening situations, such as listening to sentences in noise but can also be observed at the word level under more favorable conditions, as in one of the most commonly used tasks in audiology, i.e., repeating isolated words in silence. This study aimed to explore the role of top-down contextual influences and their dependence on lexical factors and patient-specific factors using standard clinical linguistic material. Spondaic word perception was tested in 160 hearing-impaired patients aged 23-88 years with a four-frequency average pure-tone threshold ranging from 21 to 88 dB HL. Sixty spondaic words were randomly presented at a level adjusted to correspond to a speech perception score ranging between 40 and 70% of the performance intensity function obtained using monosyllabic words. Phoneme and whole-word recognition scores were used to calculate two context-influence indices (the j factor and the ratio of word scores to phonemic scores) and were correlated with linguistic factors, such as the phonological neighborhood density and several indices of word occurrence frequencies. Contextual influence was greater for spondaic words than in similar studies using monosyllabic words, with an overall j factor of 2.07 (SD = 0.5). For both indices, context use decreased with increasing hearing loss once the average hearing loss exceeded 55 dB HL. In right-handed patients, significantly greater context influence was observed for words presented in the right ears than for words presented in the left, especially in patients with many years of education. The correlations between raw word scores (and context influence indices) and word occurrence frequencies showed a significant age-dependent effect, with a stronger correlation between perception scores and word occurrence frequencies when the occurrence frequencies were based on the years corresponding to the patients' youth, showing a "historic" word frequency effect. This effect was still observed for patients with few years of formal education, but recent occurrence frequencies based on current word exposure had a stronger influence for those patients, especially for younger ones.

Speech - Beaver Bill, 21 February 1975

A speech by Sean O'Sullivan, given in the House of Commons, "For the Recognition of the Beaver as a Symbol of the Sovereignty of the Dominion of Canada".

Development of a Biometric Personal Authentication System Based on Fingerprint and Speech

Biometrics deals with the physiological and behavioral characteristics of an individual to establish identity. Fingerprint based authentication is the most advanced biometric authentication technology. The minutiae based fingerprint identification method offer reasonable identification rate. The feature minutiae map consists of about 70-100 minutia points and matching accuracy is dropping down while the size of database is growing up. Hence it is inevitable to make the size of the fingerprint feature code to be as smaller as possible so that identification may be much easier. In this research, a novel global singularity based fingerprint representation is proposed. Fingerprint baseline, which is the line between distal and intermediate phalangeal joint line in the fingerprint, is taken as the reference line. A polygon is formed with the singularities and the fingerprint baseline. The feature vectors are the polygonal angle, sides, area, type and the ridge counts in between the singularities. 100% recognition rate is achieved in this method. The method is compared with the conventional minutiae based recognition method in terms of computation time, receiver operator characteristics (ROC) and the feature vector length. Speech is a behavioural biometric modality and can be used for identification of a speaker. In this work, MFCC of text dependant speeches are computed and clustered using k-means algorithm. A backpropagation based Artificial Neural Network is trained to identify the clustered speech code. The performance of the neural network classifier is compared with the VQ based Euclidean minimum classifier. Biometric systems that use a single modality are usually affected by problems like noisy sensor data, non-universality and/or lack of distinctiveness of the biometric trait, unacceptable error rates, and spoof attacks. Multifinger feature level fusion based fingerprint recognition is developed and the performances are measured in terms of the ROC curve. Score level fusion of fingerprint and speech based recognition system is done and 100% accuracy is achieved for a considerable range of matching threshold

Auditory word-recognition by young deaf children

This paper reviews a study of a speech discrimination test for young profoundly deaf children.

Reception of acoustic speech signals as a function of sensation levels by hearing-impaired children

This paper is a review of a study to determine the sensation level at which the best discrimination and recognition occurs for severely and profoundly deaf children and the effect of distortion (peak clipping) has on the child's ability to recognize speech.

The effect of talker age and gender on speech perception of pediatric hearing aid users

Even though pediatric hearing aid (HA) users listen most often to female talkers, clinically-used speech tests primarily consist of adult male talkers' speech. Potential effects of age and/or gender of the talker on speech perception of pediatric HA users were examined using two speech tests, hVd-vowel identification and CNC word recognition, and using speech materials spoken by four talker types (adult males, adult females, 10-12 year old girls, and 5-7 year old girls). For the nine pediatric HA users tested, word scores for the male talker's speech were higher than those for the female talkers, indicating that talker type can affect word recognition scores and that clinical tests may over-estimate everyday speech communication abilities of pediatric HA users.

Spoken word recognition in adolescents with autism spectrum disorders and specific language impairment

Spoken word recognition, during gating, appears intact in specific language impairment (SLI). This study used gating to investigate the process in adolescents with autism spectrum disorders plus language impairment (ALI). Adolescents with ALI, SLI, and typical language development (TLD), matched on nonverbal IQ listened to gated words that varied in frequency (low/high) and number of phonological onset neighbors (low/high density). Adolescents with ALI required more speech input to initially identify low-frequency words with low competitor density than those with SLI and those with TLD, who did not differ. These differences may be due to less well specified word form representations in ALI.

Effects of hearing protector devices on speech intelligibility

The purpose of this study was to determine the influence of hearing protection devices (HPDs) on the understanding of speech in young adults with normal hearing, both in a silent situation and in the presence of ambient noise. The experimental research was carried out with the following variables: five different conditions of HPD use (without protectors, with two earplugs and with two earmuffs); a type of noise (pink noise); 4 test levels (60, 70, 80 and 90 dB[A]); 6 signal/noise ratios (without noise, + 5, + 10, zero, - 5 and - 10 dB); 5 repetitions for each case, totalling 600 tests with 10 monosyllables in each one. The variable measure was the percentage of correctly heard words (monosyllabic) in the test. The results revealed that, at the lowest levels (60 and 70 dB), the protectors reduced the intelligibility of speech (compared to the tests without protectors) while, in the presence of ambient noise levels of 80 and 90 dB and unfavourable signal/noise ratios (0, -5 and -10 dB), the HPDs improved the intelligibility. A comparison of the effectiveness of earplugs versus earmuffs showed that the former offer greater efficiency in respect to the recognition of speech, providing a 30% improvement over situations in which no protection is used. As might be expected, this study confirmed that the protectors' influence on speech intelligibility is related directly to the spectral curve of the protector's attenuation. (C) 2003 Elsevier B.V. Ltd. All rights reserved.

Corpus-Based Speech Enhancement With Uncertainty Modeling and Cepstral Smoothing

We present a new approach for corpus-based speech enhancement that significantly improves over a method published by Xiao and Nickel in 2010. Corpus-based enhancement systems do not merely filter an incoming noisy signal, but resynthesize its speech content via an inventory of pre-recorded clean signals. The goal of the procedure is to perceptually improve the sound of speech signals in background noise. The proposed new method modifies Xiao's method in four significant ways. Firstly, it employs a Gaussian mixture model (GMM) instead of a vector quantizer in the phoneme recognition front-end. Secondly, the state decoding of the recognition stage is supported with an uncertainty modeling technique. With the GMM and the uncertainty modeling it is possible to eliminate the need for noise dependent system training. Thirdly, the post-processing of the original method via sinusoidal modeling is replaced with a powerful cepstral smoothing operation. And lastly, due to the improvements of these modifications, it is possible to extend the operational bandwidth of the procedure from 4 kHz to 8 kHz. The performance of the proposed method was evaluated across different noise types and different signal-to-noise ratios. The new method was able to significantly outperform traditional methods, including the one by Xiao and Nickel, in terms of PESQ scores and other objective quality measures. Results of subjective CMOS tests over a smaller set of test samples support our claims.

A speech interface for air traffic control terminals

Several issues concerning the current use of speech interfaces are discussed and the design and development of a speech interface that enables air traffic controllers to command and control their terminals by voice is presented. A special emphasis is made in the comparison between laboratory experiments and field experiments in which a set of ergonomics-related effects are detected that cannot be observed in the controlled laboratory experiments. The paper presents both objective and subjective performance obtained in field evaluation of the system with student controllers at an air traffic control (ATC) training facility. The system exhibits high word recognition test rates (0.4% error in Spanish and 1.5% in English) and low command error (6% error in Spanish and 10.6% error in English in the field tests). Subjective impression has also been positive, encouraging future development and integration phases in the Spanish ATC terminals designed by Aeropuertos Españoles y Navegación Aérea (AENA).

A Satisfaction-based Model for Affect Recognition from Conversational Features in Spoken Dialog Systems

Detecting user affect automatically during real-time conversation is the main challenge towards our greater aim of infusing social intelligence into a natural-language mixed-initiative High-Fidelity (Hi-Fi) audio control spoken dialog agent. In recent years, studies on affect detection from voice have moved on to using realistic, non-acted data, which is subtler. However, it is more challenging to perceive subtler emotions and this is demonstrated in tasks such as labelling and machine prediction. This paper attempts to address part of this challenge by considering the role of user satisfaction ratings and also conversational/dialog features in discriminating contentment and frustration, two types of emotions that are known to be prevalent within spoken human-computer interaction. However, given the laboratory constraints, users might be positively biased when rating the system, indirectly making the reliability of the satisfaction data questionable. Machine learning experiments were conducted on two datasets, users and annotators, which were then compared in order to assess the reliability of these datasets. Our results indicated that standard classifiers were significantly more successful in discriminating the abovementioned emotions and their intensities (reflected by user satisfaction ratings) from annotator data than from user data. These results corroborated that: first, satisfaction data could be used directly as an alternative target variable to model affect, and that they could be predicted exclusively by dialog features. Second, these were only true when trying to predict the abovementioned emotions using annotator?s data, suggesting that user bias does exist in a laboratory-led evaluation.

Phonotactic language recognition using i-vectors and phoneme posteriogram counts

This paper describes a novel approach to phonotactic LID, where instead of using soft-counts based on phoneme lattices, we use posteriogram to obtain n-gram counts. The high-dimensional vectors of counts are reduced to low-dimensional units for which we adapted the commonly used term i-vectors. The reduction is based on multinomial subspace modeling and is designed to work in the total-variability space. The proposed technique was tested on the NIST 2009 LRE set with better results to a system based on using soft-counts (Cavg on 30s: 3.15% vs 3.43%), and with very good results when fused with an acoustic i-vector LID system (Cavg on 30s acoustic 2.4% vs 1.25%). The proposed technique is also compared with another low dimensional projection system based on PCA. In comparison with the original soft-counts, the proposed technique provides better results, reduces the problems due to sparse counts, and avoids the process of using pruning techniques when creating the lattices.