The ITU-T Software Tool Library

The advancement of speech coding technology has led to new and improved standards. The Telecommunication Standardization Sector of the International Telecommunication Union (ITU-T) has been a leader in developing speech coding standards for telecommunications applications. To increase the efficiency of the coding standard selection process, a group called User's Group on Software Tools (UGST) was established in 1990 to produce a framework of speech coding-related software tools that would allow a common ground for activities across different organizations. Over the years, this group produced a high-quality software tool library (STL), which has been used in the definition of ITU and non-ITU standards, such as G.723.1, G.729, half-rate and enhanced full-rate GSM, among others. However, despite its usefulness, the ITU-T STL is largely unknown outside the telecommunications standards community. This paper introduces the ITU-T STL concepts and functionalities to the wider communications community.     

Listening to Natural and Synthesized Speech while Driving: Effects on User Performance

The effects of message type (navigation, E-mail, news story), voice type (text-to-speech, natural human speech), and earcon cueing (present, absent) on message comprehension and driving performance were examined. Twenty-four licensed drivers (12 under 30, 12 over 65, both equally divided by gender) participated in the experiment. They drove the UMTRI driving simulator on a road consisting of straight sections and constant radius curves, thus yielding two levels of low driving-workload. In addition, as a control condition, data were collected while participants were parked. In all conditions, participants were presented with three types of messages. Each message was immediately followed by a series of questions to assess comprehension. Navigation messages were about 4 seconds long (about 9 words). E-mail messages were about 40 seconds long (about 100 words) and news messages were about 80 seconds long (about 225 words). For all message types, comprehension of text-to-speech messages, as determined by accuracy of response to questions, and by subjective ratings, was significantly worse than comprehension of natural speech (79 versus 83 percent correct answers; 7.7/10 versus 8.6/10 subjective rating). Driving workload did not affect comprehension. Interestingly, neither the speech used (synthesized or natural) nor the message type (navigation, E-mail, news) had a significant effect on basic driving performance measured by the standard deviations of lateral lane position and steering wheel angle.     

语音识别中的加性噪声补偿研究

针对语音识别中的加性噪声进行研究，提出了动态自适应多模板谱减法和多模板谱加训练补偿法。动态自适应多模板谱减法和噪声动态自适应方法的有效结合，使谱减法既能适应环境中存在多种噪声的情况，又能有效利用当前得到的噪声环境信息。而多模板谱补偿法从谱减法的逆向角度，并基于多模式训练的思想，有效地使模板适应多种噪声情况，该方法不增加识别时的运算量，有利于考虑噪声的各种复杂情况。     

The effect of the external regulator’s absence on children’s speech use, manifested self-regulation, and task performance during learning tasks

The present study was conducted to explore the effect of the absence of the external regulators on children’s use of speech (private/social), task performance, and self-regulation during learning tasks. A novel methodology was employed through a computer-based learning environment that proposed three types/units of encouragement with only two sequences of instructional conditions, Verbal-Gesture-Silent (VGS) versus Silent-Gesture-Verbal (SGV). The Knowledge of response (KR) was applied as: verbal KR feedback with verbal encouragement during the verbal unit, visualization-representation of KR without verbal encouragement during the gesture unit, and no KR feedback without any encouragement during the silent unit. Three measurements were used: speech analysis, novel criteria to measure self-regulation and task performance, and a computer-based friendly chat questionnaire to measure children’s satisfaction. Forty preschool children were divided by their teachers between the two conditions equivalently. It was hypothesized that children in the VGS condition were more speech productive, manifested higher self-regulation, task performance, and satisfaction. The results showed significant differential effect on the speech intensity and manifested self-regulation with no significant differential effect on task performance and satisfaction during learning tasks. However, the results were not confirmed Vygotsky’s view as it were supported (neutralizing, at best) to Piaget’s view of self-regulation development.     

Auditory-phonetic projection and lexical structure in the recognition of sine-wave words.

Speech remains intelligible despite the elimination of canonical acoustic correlates of phonemes from the spectrum. A portion of this perceptual flexibility can be attributed to modulation sensitivity in the auditory-to-phonetic projection, although signal-independent properties of lexical neighborhoods also affect intelligibility in utterances composed of words. Three tests were conducted to estimate the effects of exposure to natural and sine-wave samples of speech in this kind of perceptual versatility. First, sine-wave versions of the easy and hard word sets were created, modeled on the speech samples of a single talker. The performance difference in recognition of easy and hard words was used to index the perceptual reliance on signal-independent properties of lexical contrasts. Second, several kinds of exposure produced familiarity with an aspect of sine-wave speech: (a) sine-wave sentences modeled on the same talker; (b) sine-wave sentences modeled on a different talker, to create familiarity with a sine-wave carrier; and (c) natural sentences spoken by the same talker, to create familiarity with the idiolect expressed in the sine-wave words. Recognition performance with both easy and hard sine-wave words improved after exposure only to sine-wave sentences modeled on the same talker. Third, a control test showed that signal-independent uncertainty is a plausible cause of differences in recognition of easy and hard sine-wave words. The conditions of beneficial exposure reveal the specificity of attention underlying versatility in speech perception. (PsycINFO Database Record (c) 2011 APA, all rights reserved)     

"On the relationship between lateralized brain function and orienting asymmetries": Correction to Teufel et al. (2010).

Reports an error in "On the relationship between lateralized brain function and orienting asymmetries" by Christoph Teufel, Asif A. Ghazanfar and Julia Fischer (Behavioral Neuroscience, 2010[Aug], Vol 124[4], 437-445). In the article, we wrote that “the likelihood of obtaining at least one significant result at p n. The probability of obtaining at least one significant result is therefore 1 — (1— α)n. (The following abstract of the original article appeared in record 2010-16138-001.) Hemispheric specializations for language perception constitute one of the classic topic in cognitive neuroscience. Evidence has accumulated to suggest that lateralized acoustic processing is not restricted to humans but is also found in numerous animal species. One of the methods used to track such lateralization is the orienting-asymmetry paradigm, a simple, noninvasive means to study lateralization that has been applied to a range of different species ranging from harpy eagles to humans. Here we summarize and compare the results of studies employing the orienting-asymmetry paradigm, showing that these studies yield largely inconsistent results. We critically discuss the methodology's implicit assumptions and conclude that the empirical inconsistencies produced by the orienting-asymmetry paradigm, and the lack of sufficient evidence supporting the paradigm's underlying assumptions, warrant serious caution when interpreting results obtained by the method. Nontrivial interpretations of orienting-asymmetry results will require a much better understanding of how lateralized brain functions interact with overt behaviors. (PsycINFO Database Record (c) 2011 APA, all rights reserved)     

Generalizing linguistic structures under high attention demands.

We explored whether the generalization of rules based on simple structures depends on attention. Participants were exposed to a stream of artificial words that followed a simple syllabic structure (ABA or AAB), overlaid on a sequence of familiar noises. After passively listening, participants successfully recognized the individual words present in the stream among foils, and they were able to generalize the underlying word structure to new exemplars. Yet, when attention was diverted from the speech stream (by requiring participants to monitor the sequence of noises), recognition of the individual words fell dramatically irrespective of word structure, whereas generalization depended on stimulus structure. For structures based on vowel repetitions across nonadjacent syllables (ABA; Experiment 1), generalization was affected by attention. In contrast, for structures based on adjacent repetitions (AAB; Experiment 2), generalization capacity was unaffected by attention. This pattern of results was replicated under favorable conditions for generalization, such as increased token variability and the implementation of the rule over whole syllables (Experiments 3 and 4). These results suggest a differential effect of attention on rule learning and generalization depending on stimulus structure. (PsycINFO Database Record (c) 2011 APA, all rights reserved)     

Control of interior surface materials for speech privacy in high‐speed train cabins

The effect of interior materials with various absorption coefficients on speech privacy was investigated in a 1:10 scale model of one high‐speed train cabin geometry. The speech transmission index (STI) and privacy distance (r_P) were measured in the train cabin to quantify speech privacy. Measurement cases were selected for the ceiling, sidewall, and front and back walls and were classified as high‐, medium‐ and low‐absorption coefficient cases. Interior materials with high absorption coefficients yielded a low r_P, and the ceiling had the largest impact on both the STI and r_P among the interior elements. Combinations of the three cases were measured, and the maximum reduction in r_P by the absorptive surfaces was 2.4 m, which exceeds the space between two rows of chairs in the high‐speed train. Additionally, the contribution of the interior elements to speech privacy was analyzed using recorded impulse responses and a multiple regression model for r_P using the equivalent absorption area. The analysis confirmed that the ceiling was the most important interior element for improving speech privacy. These results can be used to find the relative decrease in r_P in the acoustic design of interior materials to improve speech privacy in train cabins.     

The perception of assimilation in newly learned novel words.

The present study investigated the mechanisms underlying perceptual compensation for assimilation in novel words. During training, participants learned canonical versions of novel spoken words (e.g., decibot) presented in isolation. Following exposure to a second set of novel words the next day, participants carried out a phoneme monitoring task. Here, the novel words were presented with final alternations (e.g., decibop) in carrier sentences that either licensed assimilation (viable context: Our decibop behaved badly) or did not (unviable context: Our decibop does very well). Listeners had to monitor for the underlying form of the assimilated consonant (e.g., /t/ in decibop). Results showed more responses corresponding to the underlying form in viable than in unviable contexts. This viability effect was equivalent for novel words learned on the same day and on the previous day but was absent for unexposed control items. The processing difference between exposed and control novel words supports the idea that compensation for assimilation interacts with newly acquired phonological information and suggests that contextual compensation for assimilation is enhanced by lexical knowledge. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Externality, internality, and (in)dispensability of grammatical features in speech production: Evidence from Czech declension and conjugation.

In 3 picture–word experiments, the authors explored the activation of 2 grammatical features in Czech during lexical access: declensional class of nouns and conjugational class of verbs. Experiments 1 and 2 demonstrated congruency effects of declensional and conjugational class, respectively. Picture naming times were reliably longer if the declensional or conjugational classes of the pictures' names and the distractors were incongruent. Experiment 3 explored the origin of the congruency effect in more detail. Congruency effects were obtained for declensional class regardless of whether the target name and the distractor differed in form, speaking for competition at the lemma level. These findings are discussed in comparison with gender congruency effects. The authors propose a differentiation between externally and internally specified features of lemmas, especially with respect to the time course of their activation. Internal features that become available only when the lemma is activated (e.g., gender, declensional or conjugational class of nouns and verbs) can be bypassed or not, depending on the grammatical specification of the earlier available external features (like case or number). Following this argument, supposedly inconsistent findings regarding grammatical gender and declensional or conjugational class can be explained straightforwardly. (PsycINFO Database Record (c) 2010 APA, all rights reserved)