Valerie Hazan's home page

Valerie Hazan

Research Activities

My work is primarily concerned with phoneme categorisation, a process central to the perception of speech as it addresses the listener’s ability to organise perceived acoustic patterns into their appropriate phonemic category despite the enormous degree of inter- and intra-speaker variability in the patterns produced. More specifically, I have investigated:

• How specific acoustic patterns are used in consonant categorisation

• How phoneme categorisation develops in children
• How it can be affected by the presence of a hearing or reading impairment or the knowledge of another language
• How perception can be improved by increasing the salience of acoustic patterns
• How auditory-(visual) training can be used to improve speech perception

The impact of speaker variability on speech perception in children and adults

In a study funded by the Wellcome Trust, Duncan Markhan and I evaluated whether relative speaker intelligibility was consistent across listeners differing in age and gender. We also investigated the process of attunement to talker characteristics in children and adults and wanted to try and establish which acoustic-phonetic characteristics of the speech were correlated with intelligibility. Word intelligibility rates were obtained for 45 talkers from a homogeneous accent group from 135 listeners (adults, 11-12 year olds and 7-8 year olds) in two test conditions. Both test conditions contained multiple speakers but in the ‘single word’ condition, keywords were presented in isolation while in the ‘triplet’ condition, triplets of keywords were preceded by a precursor sentence by the same talker.  For identical word materials, word intelligibility at a signal to noise ratio of +6 dB varied significantly across talkers from 81.2 to  96.4 %. Overall, younger listeners made significantly more errors than older children or adults, and women talkers were more intelligible than other classes of talkers. The relative intelligibility of the 45 talkers was highly consistent across listener groups showing that talker intelligibility is primarily determined by talker-related factors rather than by the inter-relation of talker and listener-related factors. The presence of a precursor sentence providing indexical information did improve word intelligibility for the bottom quartile of listeners in each of the listener groups.The amount of energy present in the 1 to 3kHz energy band and speaking rate were the two factors that correlated best with intelligibility. You can find further information about the project here.

Rachel Baker and i have just begun a three-year ESRC-funded project on the impact of speaker variability on speech perception in connected discourse.

Hazan, V. and Markham, D. (2004) Acoustic-phonetic correlates of talker intelligibility for adults and children. Journal of the Acoustical Society of America, 116, 3108-3118. [link from pc with access permission to JASA]

Markham, D. and Hazan, V. (2004) The effect of talker- and listener-related factors on intelligibility for a real-word, open-set perception test. Journal of Speech, Hearing and Language Research 47, 725-737. [please contact me for reprint]

Markham, D. and Hazan, V. (2002). Speaker intelligibility of adults and children. Proceedings of International Conference for Spoken Language Processing, Denver, 16-20 September 2002, 1685-1688 [pdf]

Hazan, V. and Markham, D. (2002) Do adults and children find the same voices intelligible. ISCA Workshop on Temporal Integration in the perception of speech, P3-19.

Markham, D. and Hazan, V. (2002) The UCL Speaker Database. Speech, Hearing and Language: UCL Work in Progress, vol. 14, 1-17. [pdf]

Speech perception in children with dyslexia

An important issue in dyslexia research is whether children with reading disabilities show some deficit in speech perception, and whether this speech deficit is linked to problems with temporal processing (e.g. in their perception of rapid acoustic cues). In a study carried out in collaboration with Masterson, it was found that adults with phonological dyslexia indeed showed some weakness in speech perception, but that this was not limited to sounds that involved rapid acoustic changes (Masterson et al, 1995). A further study on speech perception abilities in children with dyslexia showed convincingly that only a subset of children showed a weakness in speech perception. Again, the problem could not solely be attributed to weakness in temporal processing but seemed more generally related to the acoustic salience of cues marking the contrasts (Adlard & Hazan, 1998). In a current Wellcome-funded project, Souhila Messaoud-Galusi, Stuart Rosen and i are investigating the speech perceptual abilities of dyslexic children in an extensive study involving 61'pure' dyslexics and 51 controls. Tests involve categorical perception tests presented in clear and in noise, natural word tests presented in noise and a battery of standardised assessment of their reading abilities and phonological awareness. Further, we are testing how within- and across-speaker variability affects speech perception in these children. Results are currently being written up for publication.

Adlard, A. and Hazan, V. (1998) Speech perception abilities in children with developmental dyslexia. Quarterly Journal of Experimental Psychology: Section A. vol 51A, 153-177. [abstract]

Masterson, J., Hazan, V. and Wiyatilake, L. (1995) Auditory discrimination problems in developmental dyslexia. Cognitive Neuropsychology,12(3), 233-259.

 

Speech perception and training in second-language learners

In an EPSRC-funded project, Marta Ortega-Llebaria, Mark Huckvale and I showed that acoustic cue-enhancement could significantly increase the perception of some phonemic contrasts by second-language learners without any training. In the following EPSRC-funded project with Andrew Faulkner and Ruth Campbell, we investigated whether visual cues (provided via video of natural face or by a talking head) could help second-language learners to acquire novel phonetic contrasts. We found that second-language learners had little sensitivity to segmental cues provided via the visual channel pre-training. Training with audiovisual materials was more successful than training with auditory materials only for the English /b/-/v/ contrast (presented to Japanese L1 speakers) which is visually-marked for native listeners but there was no difference between training modalities for the /l/-/r/ contrasts which is less clearly marked visually. In a follow-up study with Anke Sennema, we investigated whether training of the /l/-/r/ contrast with visual stimuli alone (lipreading only) could lead to successful improvements in the perception of the contrast. significant improvements were seen for novel stimuli and speakers in both visual-alone and audiovisual stimuli, following visual training. Further work on visual bias in native and non-native children and adults has been done with Chen. In a ESRC-funded project with Paul Iverson, we investigated the relative effectiveness of different techniques for speech perceptual training, and how training affected the weighting of acoustic cues for second-language learners.  

Chen, Y. and Hazan, V. (2007) Developmental Factor in Auditory-Visual Speech Perception---The McGurk Effect in Mandarin-Chinese and English Speakers. Proceedings of AVSP2007, Netherlands, 1-3 September 2007, 87-91. [paper in pdf]

Hazan, V., Sennema, A, Faulkner, A., Ortega-Llebaria, M, Iba, M. and Chung, H (2006) The use of visual cues in the perception of nonnative consonant contrasts. Journal of the Acoustical Society of America, .119, 1740-1751. [link from PC with access rights to JASA]

Hazan, V., Sennema, A., Iba, M. and Faulkner, A. (2005) Effect of audiovisual perceptual training on the perception and production of consonants in Japanese learners of English, Speech Communication, 47, 360-378. [link from PC with access rights to Speech Com]

Hazan, V. and Sennema, A. (2007) The impact of visual training on the perception and production of a non-native phonetic contrast. Proceedings of AVSP2007, Netherlands, 1-3 September 2007, 146-151. [paper in pdf]

Hazan, V. (2002) L'apprentissage des langues. Proceedings of XXIVemes Journees d'etude de la parole, Nancy, 24-27 Juin 2002, 1-5. [paper in pdf]

Iverson, P., Hazan, V, Bannister, K. (2005) Phonetic training with acoustic cue manipulations: A comparison of methods for teaching English /r/-/l/ to Japanese adults, Journal of the Acoustical Society of America, 118, 3267-3278. [link from PC with access rights to JASA]

Ortega-Llebaria, M., Faulkner, A. and Hazan, V. (2001) Auditory-visual L2 speech perception: Effects of visual cues and acoustic-phonetic context for Spanish learners of English. Proc. AVSP-2001.  [paper in pdf]

Ortega-Llebaria, M., Faulkner, A. and Hazan, V. (2001) Auditory-visual L2 speech perception: effects of visual cues and acoustic-phonetic context for Spanish learners of English. Speech, Hearing and Language: UCL Work in Progress, vol. 13, 39-41. [paper in pdf]

Ortega, M. and Hazan, V. (1999) Enhancing acoustic cues to aid L2 speech perception. Proceedings of the International Congress of Phonetic Sciences, San Francisco, 1-7 August 1999, 1, 117-120.

The development of phoneme categorisation in normally-hearing children

An  MRC-funded project investigated the development of phonemic categorisation in normally-hearing children aged 6 to 12. This work showed that speech perceptual development is a slow and gradual process as categorisation was not yet adult-like by age 12 (Hazan and Barrett, 2000). This work therefore has important implications for models of speech perception development which have mainly focused until now on the early stages of acquisition

Hazan, V. and Barrett, S. (2000) The development of phonemic categorisation in children aged 6 to 12. Journal of Phonetics, 28, 377-396. [pdf]

The development of phoneme categorisation in deaf children

In collaboration with Fourcin and Abberton, I carried out a four-year longitudinal study of the development of phoneme categorisation in a group of 16 deaf children (Hazan et al, 1991; Abberton et al.,1990). This study showed that development was much delayed relative to children with normal hearing, but that deaf children followed the same stages in the development of phoneme categorisation and the same order in the acquisition of different phonemes. This is, to my knowledge, the only study to provide a longitudinal account of the development of phoneme categorisation in deaf children. Further studies have investigated deaf children’s perception of acoustic cue information (Holden-Pitt et al., 1995) and the usefulness of phoneme categorisation as a means of speech perception assessment in comparison with more traditional speech audiometry materials (e.g. Mac Ardle et al, 2000). In collaboration with Laryngograph Ltd, the tests used in these studies have been implemented within computer-based testing software aimed at clinical users (Hazan et al, 1995).

McArdle, B., Hazan, V. and Prasher, D. (1999) A comparison of Speech Pattern audiometry and Manchester Junior Word Lists in hearing impaired children. British Journal of Audiology, 33, 383-393. [abstract]

Holden-Pitt, L., Hazan, V., Revoile, S., Edward, D., Droge, J. (1995) Temporal and spectral cue-use for initial plosive voicing perception by hearing-impaired children and normal-hearing children and adults. European Journal for Disorders of Communication, 30, 417-434. [abstract]

Hazan, V., Fourcin, A.J. and Abberton, E.R. (1991) Development of phonetic labeling in hearing- impaired children. Ear and Hearing, 12, 71-84. [abstract]

 

Development and evaluation of cue-enhanced speech

The most common approach to speech enhancement is to use signal processing techniques to ‘clean up’ noisy signals. Our ‘phonetic’ approach to speech enhancement, supported by two successive EPSRC grants, is quite different as it aimed to make speech more resistant to degradation by selectively amplifying perceptually-important regions of the signal before it is degraded by noise. Significant improvements in intelligibility were obtained for cue-enhanced nonsense words and sentences presented in noise (Hazan & Simpson, 1998). Cue-enhancement was also shown to significantly improve speech intelligibility for non-native speakers, who tend to be particularly affected by noisy environments (Hazan & Simpson, 2000). The UCL Enhance software, produced by Dr Mark Huckvale, enables users to enhance signals using our approach or more standard techniques. It is freely available at the UCL Enhance website.

Hazan, V. and Simpson, A. (2000). The effect of cue-enhancement on consonant intelligibility in noise: speaker and listener effects. Language and Speech, 43, 273-294.

Hazan, V. and Simpson, A. (1998) The effect of cue-enhancement on the intelligibility of nonsense word and sentence materials presented in noise. Speech Communication, 24, 211-226.

Development of assessment methods for speech output systems

As part of the ESPRIT-SAM project, I was involved in the development of semantically-unpredictable (SUS) sentence material for use in the assessment of speech output systems. The SUSgen package which allows for the automatic generation of SUS material in 6 European languages is available in DOS format. Contact me if you would like a copy.

Benoit, C., Grice, M. & Hazan, V. (1996) The SUS test: a method for the assessment of text to speech synthesis intelligibility using Semantically Unpredictable Sentences. Speech Communication, 18, 381-392. [pdf]