Roles and representations of systematic fine phonetic detail in speech understanding

Hawkins, Sarah

doi:10.1016/j.wocn.2003.09.006

Cited by 220 publications

(180 citation statements)

References 77 publications

Supporting

Mentioning

164

Contrasting

Unclassified

Order By: Relevance

“…It is either an explicit or implicit presumption of most models of spoken-language perception that linguistic processing begins with discrete phoneme representations, isolated from the messy acoustics involved in their characterization. However, research has undermined this view, as effects of acoustic variation are evident in word recognition (e.g., Hawkins, 2003) and expectations from word and semantic representations feed back to influence speech-sound perception (McClelland, Mirman, & Holt, 2006). Indeed, it is difficult to determine where auditory perception ends and cognition begins; the perception/cognition boundary may be artificial, and assuming its existence may be counterproductive to progress in understanding.…”

Section: An Auditory Cognitive Sciencementioning

confidence: 99%

Speech Perception Within an Auditory Cognitive Science Framework

Holt

Lotto

2008

Curr Dir Psychol Sci

View full text Add to dashboard Cite

The complexities of the acoustic speech signal pose many significant challenges for listeners. Although perceiving speech begins with auditory processing, investigation of speech perception has progressed mostly independently of study of the auditory system. Nevertheless, a growing body of evidence demonstrates that cross-fertilization between the two areas of research can be productive. We briefly describe research bridging the study of general auditory processing and speech perception, showing that the latter is constrained and influenced by operating characteristics of the auditory system and that our understanding of the processes involved in speech perception is enhanced by study within a more general framework. The disconnect between the two areas of research has stunted the development of a truly interdisciplinary science, but there is an opportunity for great strides in understanding with the development of an integrated field of auditory cognitive science.Keywords speech perception; auditory perception; auditory cognitive scienceThe ease with which a listener perceives speech in his or her native language belies the complexity of the task. A spoken word exists as a fleeting fluctuation of air molecules for a mere fraction of a second, but listeners are usually able to extract the intended message. The seemingly trivial ability to determine that the spoken words dean, den, dune, and dawn begin with the same English consonant, /d/, is actually a remarkable accomplishment. There are a number of physical acoustic characteristics associated with the production of /d/, and they unfold quickly across just tens of milliseconds. These acoustic correlates vary with the following vowel (such that the initial sound in dean is distinct from that of dune) and the preceding word context, as well as with the dialect, gender, emotional state, and physical stature of the speaker. To make matters more challenging, the particular acoustic correlates utilized by a listener depend on that listener's native language and expectations. Invariant perception ("these are all /d/s") in the face of variable acoustic signals has been one of the central puzzles in the study of speech perception for more than 50 years.Presumably, understanding how the auditory system processes complex sounds can be informative about how speech perception is accomplished; and, likewise, investigation of speech perception can provide clues to how the auditory system functions. Unfortunately,

show abstract

“…The data are compatible with this category retuning account, as long as it is modified so that perceptual learning is not based solely on adjustments to phonemic representations, but can also involve aspects of syllabic and prosodic structure (Hawkins & Smith 2001;Hawkins 2003;Hawkins 2010;Pierrehumbert 2002Pierrehumbert , 2006. By this account, listeners would not only store phonemic categories and learn about speaker-specific realisation of these (Norris et al 2003), but would additionally store information about prosodic structures and processes, and learn about -52 -speaker-specific phonetic implementation of these.…”

Section: Discussionmentioning

confidence: 55%

Production and perception of speaker-specific phonetic detail at word boundaries

Smith

Hawkins

2012

Journal of Phonetics

Self Cite

View full text Add to dashboard Cite

Experiments show that learning about familiar voices affects speech processing in many tasks. However, most studies focus on isolated phonemes or words and do not explore which phonetic properties are learned about or retained in memory. This work investigated inter-speaker phonetic variation involving word boundaries, and its perceptual consequences. A production experiment found significant variation in the extent to which speakers used a number of acoustic properties to distinguish junctural minimal pairs e.g. So he diced them-So he'd iced them. A perception experiment then tested intelligibility in noise of the junctural minimal pairs before and after familiarisation with a particular voice. Subjects who heard the same voice during testing as during the familiarisation period showed significantly more improvement in identification of words and syllable constituents around word boundaries than those who heard different voices. These data support the view that perceptual learning about the particular pronunciations associated with individual speakers helps listeners to identify syllabic structure and the location of word boundaries.

show abstract

“…Goldinger, 1998;Johnson, 2006;Lachs, McMichael, & Pisoni, 2003;Hawkins, 2003;Pierrehumbert, 2006). SSBE speech elicited greater task improvement after familiarization than GE speech, while GE participants also showed more improvement.…”

Section: Interaction Of Long-term Familiarity and Short-term Familiarmentioning

confidence: 80%

Cross-Accent Intelligibility of Speech in Noise: Long-Term Familiarity and Short-Term Familiarization

Smith

Holmes-Elliott

Pettinato

et al. 2014

Quarterly Journal of Experimental Psychology

View full text Add to dashboard Cite

Listeners must cope with a great deal of variability in the speech signal, and thus theories of speech perception must also account for variability, which comes from a number of sources, including variation between accents. It is well-known that there is a processing cost when listening to speech in an accent other than one's own, but recent work has suggested that this cost is reduced when listening to a familiar accent widely represented in the media, and/or when short amounts of exposure to an accent are provided. Little is known, however, about how these factors (long-term familiarity and short-term familiarization with an accent) interact. The current study tested this interaction by playing listeners difficult-to-segment sentences in noise, before and after a familiarization period where the same sentences were heard in the clear, allowing us to manipulate shortterm familiarization. Listeners were speakers of either Glasgow English or Standard Southern British English, and they listened to speech in either their own or the other accent, thereby allowing us to manipulate long-term familiarity. Results suggest that both long-term familiarity and short-term familiarization mitigate the perceptual processing costs of listening to an accent that is not one's own, but seem not to compensate for them entirely, even when the accent is widely heard in the media.3

show abstract

“…Of course, the different scientific communities that study human spoken language and speech technology systems are not without their own ideas about where future progress might lie in their respective niche areas (Greenberg, 1996;Cooke, 2003;Hawkins, 2003;Lee, 2004;Morgan et al, 2005;Moore, 2005c). However, what is missing is a truly integrated view that not only To appear in: Speech Communication, Special Issue on Bridging the Gap Between Human and Automatic Speech Processing, (2007) Page 4 of 47 draws the relevant pieces of knowledge together, but which also serves to provide a coherent explanation of what is, after all, a single behaviour.…”

Section: The Scale Of the Problemmentioning

confidence: 99%

Spoken language processing: Piecing together the puzzle

Moore

2007

Speech Communication

View full text Add to dashboard Cite

Attempting to understand the fundamental mechanisms underlying spoken language processing, whether it is viewed as behaviour exhibited by human beings or as a faculty simulated by machines, is one of the greatest scientific challenges of our age. Despite tremendous achievements over the past fifty or so years, there is still a long way to go before we reach a comprehensive explanation of human spoken language behaviour and can create a technology with performance approaching or exceeding that of a human being. It is argued that progress is hampered by the fragmentation of the field across many different disciplines, coupled with a failure to create an integrated view of the fundamental mechanisms that underpin one organism's ability to communicate with another. This paper weaves together accounts from a wide variety of different disciplines concerned with the behaviour of living systems -many of them outside the normal realms of spoken language -and compiles them into a new model: PRESENCE (PREdictive SENsorimotor Control and Emulation). It is hoped that the results of this research will provide a sufficient glimpse into the future to give breath to a new generation of research into spoken language processing by mind or machine.

show abstract

“…Kemps et al (2005) found that, in addition to duration, intonation seems to help the perceptual system in distinguishing singular forms from the stems of plural forms. However, currently no computational models of HSR exist that are able to model this fine phonetic variation (Hawkins, 2003). Our ultimate goal is to refine an existing computational model of HSR 'SpeM' (Scharenborg et al, 2005) such that it is able to capture and use fine-grained acousticphonetic variation during speech recognition.…”

Section: Introductionmentioning

confidence: 99%

Towards capturing fine phonetic variation in speech using articulatory features

Scharenborg

Wan

Moore

2007

Speech Communication

View full text Add to dashboard Cite

The ultimate goal of our research is to develop a computational model of human speech recognition that is able to capture the effects of fine-grained acoustic variation on speech recognition behaviour. As part of this work we are investigating automatic feature classifiers that are able to create reliable and accurate transcriptions of the articulatory behaviour encoded in the acoustic speech signal. In the experiments reported here, we analysed the classification results from support vector machines (SVMs) and multilayer perceptrons (MLPs). MLPs have been widely and successfully used for the task of multi-value articulatory feature classification, while (to the best of our knowledge) SVMs have not. This paper compares the performance of the two classifiers and analyses the results in order to better understand the articulatory representations. It was found that the SVMs outperformed the MLPs for five out of the seven articulatory feature classes we investigated while using only 8.8–44.2% of the training material used for training the MLPs. The structure in the misclassifications of the SVMs and MLPs suggested that there might be a mismatch between the characteristics of the classification systems and the characteristics of the description of the AF values themselves. The analyses showed that some of the misclassified features are inherently confusable given the acoustic space. We concluded that in order to come to a feature set that can be used for a reliable and accurate automatic description of the speech signal; it could be beneficial to move away from quantised representations

show abstract

Roles and representations of systematic fine phonetic detail in speech understanding

Cited by 220 publications

References 77 publications

Speech Perception Within an Auditory Cognitive Science Framework

Production and perception of speaker-specific phonetic detail at word boundaries

Cross-Accent Intelligibility of Speech in Noise: Long-Term Familiarity and Short-Term Familiarization

Spoken language processing: Piecing together the puzzle

Towards capturing fine phonetic variation in speech using articulatory features

Contact Info

Product

Resources

About