CkBooks Online – Free Online Books

PATTERN RECOGNITION (continued)

Voice-Onset Time

Voice is human sound and onset means beginning. In the pronunciation of such consonants as

[b] and [p], two things happen: Closed lips are opened, releasing air and the vocal cords begin vibrating (voicing). In voiced consonants such as [b], the release of air and voicing is nearly simultaneous. In unvoiced consonants such as [p], the release occurs 60 ms before the vibration begins. What we are detecting when we perceive a voiced consonant versus an unvoiced consonant is the presence or absence of a 60 ms interval between release and voicing. This period of time is called Voice-Onset Time.

And when release and voicing come at the same time then it is called voiced sound like ba.

An experiment was conducted to identify the voice-onset time. They presented two words anbil and anpil. They wanted to know the voice timing difference between b and p. The diagram is showing results. In Sound wave the first arrow is showing release. Second arrow is showing voicing. And the difference is 60 ms. P has long interval between release and voicing and b has short interval.

Adaptation paradigm

Eimas & Corbit (1973) conducted an experiment that called adaptation paradigm. When we listen a sound again and again we expect this sound will not come again. This is called the fatigue paradigm or adaptation paradigm.

Eimas and Corbit had their subjects listen to repeated presentations of da (voiced). This sound involves a voiced consonant, [d]. They then presented objects with a series of artificial sounds that spanned the acoustic continuum such as that between ba (voiced) and pa (voiceless). Subjects had to indicate whether each of these artificial stimuli sounded more like ba or more pa. They found that subjects who under normal conditions would report ba as ba were now reporting it as pa. They found that subjects who under normal conditions would report ba as ba were now reporting it as pa. They reasoned that repeated voicing makes the perceptual system adapt to voicing and expect unvoiced stimuli. This experiment is very important because it tells us not the sound but itself features are being detected. That is a critical thing about the series of experiment.

Basic features of feature analysis model;

Cognitive psychology is trying to find the causes of these types of mistakes. Because usually behind these mistakes, is a theory or model.

Whenever we can not see the problems we can not find the weakness of systems. So we can not explain any system without understanding its process.

In model of hearing we talk about template matching. There are thousands of sounds in the world. Like different sounds make by tonga wala have meanings in other languages. Sounds are different in different languages. A small feature can change all meanings of sentence or words. Some words vary with tone.

Cognitive psychology is not just about laboratory experiments. It is about our daily life. Speech recognition is also related with its features.