Digital Signal
Processing
Project On Speech Recognigation Using MATLAB
Submitted By Submitted To
Ehsanul Haque
1. Rabin
Id: 19202208007
Asiful Hoque
Lecturer
29th Intake Sec:01 Department Of Electrical &
Abid Hasan
2. Id: 19202208023
Electronic Engineering
Bangladesh University Of Business & Technology
29th Intake Sec:01
Md.Rajon Ali
3. Id: 19202208033
29th Intake Sec:01
Mehedi Hasan
4. Id: 19202208034
29th Intake Sec:01
Main Contents
Methodology Simulation
Introductory Explanation about the Coding Simulate the code &
theoritical Show the Output
Basic details About Written Project
Implemantion
our Project. Commanding code
Introductory
Introductio
1. Introduce about the project
Introductory Parts
2. Objective
Discuss about the objectives of
our project.
Litareture
3. Review
Know about the history & other
related projects.
Problem
4. Statement
Discuss about the disadvantages
of our project
Introduction
In the modern world, there is an ever-increasing need to authenticate and identify
individuals automatically. Securing personal privacy and deterring identity theft are
national priorities.
Such on demand of security purpose we can use voice recognition to access our
private data.
Objective
This project tried to recognize the speach identification when he/she speaks with others. For smooth and sophisticated SR, the Matlab digital
signal-processing tools box and programmable functions are very helpful.
Speakers may input their spoken language through the Matlab workstation by recording it in the digital format such as WAV or MP3. The
channel type such as single channel or double channel may also be chosen with the speech rate for the best signal presentation.
A reference signal can be saved, which may be chosen for reference while performing the comparison on further stages:
• The frequency is recorded to the data base.
• Sample voice record & Digital filters must be utilized for removing noise components in the signal.
Using the correlation method speach matching rate analysis.
• After the signal settlement, we compared it with the reference database for evaluating the
characteristics of the speaker.
Litareture Review
The idea of speech recognition was presented in 1940s but the Bell Labs presented the first speech
recognition system in 1952, which could only perform to some extent in complete silence. So, the
foundations of speech recognition were laid in 40s and 50s because the baseline achievements were
accomplished during those two decades including presenting information and automation theoretical models.
During the 1960s, the speech recognition systems were improved to the extent that they could recognize
small vocabulary (10 to 100 words) using basic acoustic-phonetic speech sound properties. It was an
important decade because scientists and linguists joined hands to develop time normalization methods and
filter banks.
Types of speech recognition
*Isolated Speech
*Connected Speech
*Continuous speech
*Spontaneous Speech
SWOT analysis
Strengths Weaknesses
Earth is the third planet from
the Sun and the only one that
S W Neptune is also the fourth
largest planet by diameter in
harbors life the Solar System
Opportunities Threats
Mercury is the closest planet
to the Sun and the smallest of
O T Mars is full of iron oxide
dust, which gives the planet
them all its reddish cast
Problem Statement
Problem 1 Problem 2 Problem 3 Week 4
Record the voice in Correlation value Neptune is the Jupiter is the biggest
correct format matching farthest planet from planet in the Solar
the Sun System
Methodology
Project overview
A speech recognition process should be implemented using Matlab to provide
speaker identification services. This system is viable for the security, and it may
be used for preventing un-authorized access into a particular system/data.
Advance digital signal processing designs a Speech Recognition Algorithm
(SRA)using auto correlation. Accurate and practical speech recognition results in
smooth and quick recognition.
Methodology
Project overview
Cross Correlation (CC)
Assuming the recorded speech signal of a similar/same word, and two speech signal spectrums are also similar.
When cross-correlation is calculated for two same spectrums, and the results are plotted, the cross-correlation is
symmetric by definition. After finding crosscorrelation, the frequency spectrums of a couple of recorded speech
signal, so we need to compare them.
Autocorrelation (AC)
It can be found by computing the cross-correlation of a particular signal and itself. It will not correlate with the
remaining/other signals. Autocorrelation algorithms measure the extent to a signal's self-correlation. Thus, a training
signal's autocorrelation comparison finds out the least differences among autocorrelations.
Methodology
Databaseinformation
M
D
Auto correlation A I
S
T
P
C L
Pitch detector A
H Y
Microphone signal information
Project Explanation
Speech recording
A signal is shifted to Matlab through voice recording with a microphone. Generally, highquality
recording is required for appropriately following many norms needed for the required quality.
The channels of recording may be single or double (stereo/mono), and the voice signal
frequency is normally 44,100 Hz and 16-bit resolution is needed for recording voice/sound
signal. Figure 4.1 depicts the data input to Matlab.
Speaker input .WAV after receiving the input signal from microphone
Database Signal sampling
Signal slotting @3 second
Unvoiced part filtering
Plotting the result
Auto correlation
Pitch detection
Matching filter
Results
Thanks!
Do you have any questions?
[email protected]
+91 620 421 838 | yourwebsite.com
Credits: This presentation template was created by
Slidesgo, including icons by Flaticon and infographics &
images by Freepik
Please keep this slide for attribution
Icon pack
Alternative resources
Here’s an assortment of alternative resources whose style fits the one of this template:
Vectors
● White background with blue circuits
Resources
Did you like the resources on this template? Get Photos
them for free at our other websites: ● Modern and equipped computer lab
● Medium shot programmer sitting on chair
Vectors ● Rpa concept with hand touching screen
● Flat technology cyber monday background ● Focused man working on computer
● Medium shot smiley man talking on phone
Icons ● Smiley woman working with laptop
● Programming icon pack | Filled ● Low angle businessman
Instructions for use (premium users)
As a Premium user, you can use this template without attributing Slidesgo or keeping the "Thanks" slide.
You are allowed to:
● Modify this template.
● Use it for both personal and commercial purposes.
● Hide or delete the “Thanks” slide and the mention to Slidesgo in the credits.
● Share this template in an editable format with people who are not part of your team.
You are not allowed to:
● Sublicense, sell or rent this Slidesgo Template (or a modified version of this Slidesgo Template).
● Distribute this Slidesgo Template (or a modified version of this Slidesgo Template) or include it in a database or in
any other product or service that offers downloadable images, icons or presentations that may be subject to
distribution or resale.
● Use any of the elements that are part of this Slidesgo Template in an isolated and separated way from this
Template.
● Register any of the elements that are part of this template as a trademark or logo, or register it as a work in an
intellectual property registry or similar.
For more information about editing slides, please read our FAQs or visit Slidesgo School:
https://slidesgo.com/faqs and https://slidesgo.com/slidesgo-school
Fonts & colors used
This presentation has been made using the following fonts:
Exo
(https://fonts.google.com/specimen/Exo)
PT Sans
(https://fonts.google.com/specimen/PT+Sans)
#ffffff #13005b #ffc800 #11cfe7 #f788f7
Storyset
Create your Story with our illustrated concepts. Choose the style you like the most, edit its colors, pick
the background and layers you want to show and bring them to life with the animator panel! It will
boost your presentation. Check out how it works.
Pana Amico Bro Rafiki Cuate
Use our editable graphic resources...
You can easily resize these resources without losing quality. To change the color, just ungroup the resource and click
on the object you want to change. Then, click on the paint bucket and select the color you want. Group the resource
again when you’re done. You can also look for more infographics on Slidesgo.
JANUARY FEBRUARY MARCH APRIL MAY JUNE
PHASE 1
Task 1
Task 2
PHASE 2
Task 1
Task 2
JANUARY FEBRUARY MARCH APRIL
PHASE 1
Task 1
Task 2
...and our sets of editable icons
You can resize these icons without losing quality.
You can change the stroke and fill color; just select the icon and click on the paint bucket/pen.
In Google Slides, you can also use Flaticon’s extension, allowing you to customize and add even more icons.
Educational Icons Medical Icons
Business Icons Teamwork Icons
Help & Support Icons Avatar Icons
Creative Process Icons Performing Arts Icons
Nature Icons
SEO & Marketing Icons
Premium infographics
Text 1 Text 2 Text 3
“Mercury is the “Despite being red,
smallest planet in Mars is actually a
the Solar System” very cold place”
Add the title here 2016 2017 2018 2019 2020
“Venus is the
“Neptune is the “Saturn is a gas
second planet from
farthest planet from giant and has
the Sun and is
the Sun” several rings”
terribly hot”
Text 4 Text 5 Text 6
Premium infographics
75% “Despite being red,
Mars is actually a
very cold place”
75%
“Jupiter is the
TITLE 1 biggest planet in the
entire Solar System”
“Mercury is the
smallest planet in the
Solar System”
“Saturn is a gas
giant and has
several rings”
Premium Icons
Digital Marketing
Premium Icons
Online Learning
Premium Icons
Laboratory
Premium Icons
Goals & Results
Premium Icons
Infographic Elements
