An Introduction to Speech Recognition - Binghamton University

movedearΤεχνίτη Νοημοσύνη και Ρομποτική

17 Νοε 2013 (πριν από 3 χρόνια και 6 μήνες)

72 εμφανίσεις

Speech
Recognition
Speech
Recognition

Hongbing Hu
Department of Electrical and Computer
Engineering, Binghamton University
04/29/2008
1
Electrical and Computer Engineering
Binghamton University, State University of New York
Electrical and Computer Engineering
Binghamton University, State University of New York
SpeechRecognitionArchitecture
Speech

Recognition

Architecture
SpeechWaveform
Speech

Waveform
Feature
Extraction
Speech Features
Recognizer (HMM/NN)
Classification
(Recognition)
ini:dsil
Phonemes
2
Electrical and Computer Engineering
Binghamton University, State University of New York
Electrical and Computer Engineering
Binghamton University, State University of New York
I need a
Words
Transformation
Transformation
5000
6000
5000
6000
Female Speaker
Male Speaker
e
ncy
4000
e
ncy
3000
4000
Sh e
Sh e
Frequ
e
2000
3000
Frequ
e
2000
3000
1000
0
1000
3
Electrical and Computer Engineering
Binghamton University, State University of New York
Electrical and Computer Engineering
Binghamton University, State University of New York
Time
0
0.2
0.4
0.6
0.8
Time
0.20.40.60.8
0
RecognitionbasedonHMMs
Recognition

based

on

HMMs
8000
10000
12000
2000
4000
6000
Frequency (Hz)
4000
6000
8000
10000
Frequency (Hz)
Sh e h a s ..
Sh e h a s ..
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
0
2000
Time
0.1
0.2
0.3
0.4
0.5
0
2000
Time

Hidden Markov Models (HMMs)
S
i
: State
S1
Start
End
S2
S3
i
Sh e h a s ..
4
Electrical and Computer Engineering
Binghamton University, State University of New York
Electrical and Computer Engineering
Binghamton University, State University of New York
LanguageProcessing
Language

Processing
4000
6000
8000
0.2
0.4
0.6
0.8
1
1.2
1.4
0
2000
Sh
e
sil
has
sil
a
sil
Phoneme
Sh
e

sil
h

a

s

sil
a
sil
..
She has a
g
oo
d
Phoneme
Word
g
Shh
iildhlh
Word
St
Sh
e
h
as an
i
ce sm
il
e, an
d

h
e a
l
so
h
as …
S
en
t
ence
5
Electrical and Computer Engineering
Binghamton University, State University of New York
Electrical and Computer Engineering
Binghamton University, State University of New York
App1:SpeechUserInterfaceforComputers
App
.
1:

Speech

User

Interface

for

Computers

Speech Recognition in Windows Vista
C

C
ommanding

Dictation (Typing)




ﱰ







ﱰ


Support for multiple languages
6
Electrical and Computer Engineering
Binghamton University, State University of New York
Electrical and Computer Engineering
Binghamton University, State University of New York
App2:SpeechRecognitionforRobots
App
.
2:

Speech

Recognition

for

Robots

AIBO: Robotic Pet

Voice Recognition

Record Name

北ﵩョ﹣ョ



省若說

﹣ョ


Reception Robot for Hospitals
7
Electrical and Computer Engineering
Binghamton University, State University of New York
Electrical and Computer Engineering
Binghamton University, State University of New York
App3:SpeechRecognitioninVehicles
App
.
3:

Speech

Recognition

in

Vehicles

SYNC System (Ford and Microsoft)

Voice-activated destination

Voice-activated audio



ﱩﵡョャ




省

ﹴ


Electrical and Computer Engineering
Binghamton University, State University of New York
Electrical and Computer Engineering
Binghamton University, State University of New York