2.04 - Visual Aspects of Speech & Speech Reading Flashcards
Are the terms “Lipreading” vs. “Speechreading” used interchangeably by some in AR field?
Yes
What is the difference between Lipreading vs. Speechreading?
Lipreading: The process of using visual signal for recognizing speech
Speechreading: speech recognition using both auditory and visual cues
What is the process of using visual signals for recognizing speech: Lipreading vs. Speechreading?
Lipreading
What is the process of using only visual cues provided by talker’s face: Lipreading vs. Speechreading?
Lipreading
What is the process of speech recognition using both auditory and visual cues: Lipreading vs. Speechreading?
Speechreading
What uses any available cues: the speaker’s facial expressions and gestures, body postures, contextual cues: Lipreading vs. Speechreading?
Speechreading
What integrates what is heard with what is seen: Lipreading vs. Speechreading?
Speechreading
Which is Analytical: Lipreading vs. Speechreading?
Lipreading
Which is Synthetic: Lipreading vs. Speechreading?
Speechreading
What are the five factors that influence Lipreading?
- Visability
- Rapidity
- Coarticulation & Stress Effects
- Visemes & Homophenes (Sparse or dense lexical neighborhood)
- Talker Effects
____% of speech sounds are not visible on the mouth. Most people recognize ____% of words they see.
60%
20%
What places/types of articulation are the MOST visable?
3
Bilabials
Labiodentals
Interdentals
What places/types of articulation are the LEAST visable?
5
Voiced
Alveolars
Palatals
Velars
Vowels
What phonemes are MOST visible?
10
/p/, /b/, & /m/
/w/ & /ʍ/
/f/ & /v/
/ʃ/ & /ʧ/
/θ/
What phonemes are LEAST visible?
5
/k/ & /g/
/t/
/s/ & /z/
Rapidity:
Average speaker rate: ____ phonemes/s
15
Visual resolution
The eye is only capable of distinguishing _____ discrete mouth
movements/second.
9-10
Are word boundaries clear visually?
No
Difficult to tell where one word ends and another starts
Coarticulation & Stress
Can the same sound look different depending on its phonetic and linguistic context?
Yes
/b/ in beet looks different than in boot (coarticulation effect)
Talker’s stress pattern can change appearance of words (“How ARE you?” versus “How’re you?”)
What are Visemes?
A group of speech sounds that look identical on the lips
Example: /b/, /m/, & /p/
What are Homophenes?
Groups of words that look the same on the mouth
What percentage of English words are homophenes?
40‐60%
What are some Talker Effects that can affect intelligibility?
(3)
Regional accents
Mouth/Lip movement
Expressiveness
Does Speechreading use both auditory and visual cues?
Yes
What kind of information does auditory signals give?
Voicing (voice vs. voiceless)
Envelope (time‐intensity variations of the audio signal produced during speech)
Manner (stops, fricatives, nasals, affricates, glides)
What kind of cues are provided by visual information?
4
Place of articulation
Emotions
Message type
Etc.
What is Audio‐Visual Integration?
When information from the auditory and the visual signal combine to form a unified percept