Please enable JavaScript.
Coggle requires JavaScript to display documents.
Speech Recognition (History (1952 (1st system) "Audrey" by Bell…
Speech Recognition
Advantages
Dictation rather than writing
Good Consumer experience
Less use of input devices
time saver
Steps
Process
Hidden Markov Modeling process
Goal:
Finding hidden states
Analyzes sequential data
Recognizes word in an unknown file
Output
Recognized Speech
Input
Audio Signal (in folder)
Challenges
silent, quite environment needed
User's vocal strain
May not understand homonyms
Might not understand if the user signal through the microphone is unclear
Popular apps/systems
Siri
Cortana
Bing
Alexa
History
1952 (1st system) "Audrey" by Bell Laboratories
1962 IBM's Shoebox
1971-1976 (Research in this field)
1970's - Improvements in the sector "
DARPA
"
1980's - Word count increased from a few hundreds to many thousands
Future Scope
Portable hand-held devices for speech recognition dictation
Greater scope of understanding speech intentions than pure meaning
Better accuracy
Assists in other AI related developments