Top Banner
Music recognition searching application Member A:Kwan Lok Hei Aaron (12213802) Member B:XIE Siyu Shelley (12250643) Member C:Cheung Lik Man(12206970) 3/24/22
37
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: Music recognition

Music recognition searching application

Member A:Kwan Lok Hei Aaron (12213802)

Member B:XIE Siyu Shelley (12250643)

Member C:Cheung Lik Man(12206970)

Saturday, April 15, 2023

Page 2: Music recognition

WE WANT TO KNOW…..

Page 3: Music recognition

Can music recognition

searching applications be

more popular?

Page 4: Music recognition

Are the music recognition searching

applications really helpful?

Page 5: Music recognition

Objectives

Briefly introduce music recognition

Compare the music recognition searching engine with ordinary searching engine

The Merits of music recognition searching

The limitations of music recognition searching

Find out what does the users really need for the music searching engine

Page 6: Music recognition

Music recognition searching application

It is one of the applications using Speech recognition technology.

The application relies on: Fingerprinting music (spectrogram) technology.

Speech-to-text technology.

Operation:

1. Abundant of music’s fingerprints are stored in the database.

2. A user can tape the song for 10 seconds in order to get sufficient information(fingerprint).

3. The app uploads the fingerprint to its corresponding service and do a matching.

4. If the match is found, the app outputs the result to the user.

Page 7: Music recognition

Application(1) Features:

Biographies

Other albums and tracks by that artist

Lyrics display

Show the YouTube clips

Show similar artists

Tour dates

Song’s album appearance

Share via Email, Twitter, Facebook or SMS

Page 8: Music recognition

Application(2)

Features:

Artist biographies

Tour dates

A map of the location where you tagged the song

Show discography of the artist

Lyrics display

Offer YouTube clips

Recommended songs relating to the one you tagged

Share via Twitter and Facebook

Shazam

Page 9: Music recognition

Video demonstration(Shazam1)

https://www.youtube.com/watch?feature=player_embedded&v=hny-G-0nUBM

Page 10: Music recognition

Video demonstration(Shazam2)

https://www.youtube.com/watch?feature=player_embedded&v=BJzCyaE6eKQ

Page 11: Music recognition

Video demonstration(Soundhound)

https://www.youtube.com/watch?v=7c1MnRaiRwg

Page 12: Music recognition

Job Division

Survey:

Member A,B,C

Question:11-13

Member A

Analysis Chart Creation

Question:3-5

Member B

Analysis Chart Creation

Question:6-10

Member C

Analysis Chart Creation

Data Process:

Member A,B,C

PowerPoint Creation:

Member C

Information collection: Member A

Mind Map:

Member A,B,C

Web Search:

Member B

Page 13: Music recognition

Information Collected

Web Searching

Searching tools: Google

Keywords for searching:

“voice search” technology

Voice OR speech recognition principle

Voice recognition –history

Speech recognition *

Voice recognition site: http://library.hkbu.edu.hk/main/index.html

Page 14: Music recognition

Information Collected(2)

Library Resources:

Onesearch

Keywords for searching:

“Music recognition”

Shazam OR Soundhound

Page 15: Music recognition

Mind Map

http://www.mindmeister.com/251475311/voice-recognition

Page 16: Music recognition

Online Survey

Aim: To know what does the users really need for the

music searching engine

To know what is the advantages of music recognition searching application over ordinary searching engine

Survey Link: https://qtrial.qualtrics.com/SE/?SID=SV_55XIo6wFYrlinbL

No. of respond: 33

Page 17: Music recognition

3.How much do you like listening music?

Degree Nauseous

Tiresome

Neutral Favorite Very like

Total Response

Answer 0 0 6 12 15 33

Page 18: Music recognition

4.What is your frequency of listening music in a month?

Frequency Response %

Seldom 0 0%

1-2 time weekly 3 9%

2-5 times weekly 5 15%

5-9 times weekly 10 30%

Above 9 times weekly

15 45%

Total 33 100%

Page 19: Music recognition

MP3

Cell phone

Computer

Other devices

0

1

2

3

4

5

6

7

8

Frenquency and Devices of Listening to Music

1-2 time weekly

2-5 times weekly

5-9 times weekly

Above 9 times weekly

Page 20: Music recognition

5.What do you often use to listen to music?

Device Response %

MP3 4 12%

Cell phone 23 70%

Computer 20 61%

Other devices 4 12%

Page 21: Music recognition

Analysis to Question 3-5

2 Factors that reduce music recognition

apps’ popularity:

1. Degree of Likeness

2. Advantages of Other Devices

Page 22: Music recognition

19%

37%

27%

4%13%

Q6.What do you use when searching for a piece of music?

A sentence of the lyricsThe name of a songThe singerThe melodyAsking to friends

Page 23: Music recognition

The best way

A better way

Normal way

A worse way

The worst way

0 5 10 15 20 25 30 35

7

14

8

4

0

21

11

1

0

0

4

7

13

9

0

1

1

11

19

1

0

0

0

1

32

10.What tools do you think that is the best way to search unknown music or songs?(please arrange in order)

OtherGoing to CD shopAsking friendsInternetMusic recognition applica-tions or software

Page 24: Music recognition

Analysis(1)

Most people always need to remember the detail of the music or song such as the name, the lyrics and the singer in order to find the music or song that they desire to listen.

They seldom use the melody to find the music or song.

Although most of the respondents chose Internet as the best way for searching music or song, music recognition application or software is the second popular one.

Music recognition application or software is getting people’s awareness and support.

Page 25: Music recognition

YESNO

0

5

10

15

20

25

30

35

Q7.If an app can help you searching for music efficiently when you don’t know much about the song, will you use it?

Series1

Page 26: Music recognition

Strongly Disagree

Disagree

Neither Agree nor Disagree

Agree

Strongly Agree

0 2 4 6 8 10 12 14

Q9.Do you think such apps can replace the music searching website?(eg. google, yahoo, youtube)

Strongly DisagreeDisagreeNeither Agree nor DisagreeAgreeStrongly Agree

Page 27: Music recognition

Analysis(2)

The demand and the popularity of music recognition application or software is being increase.

People think this innovative recognition technology may replace the traditional searching methods.

Page 28: Music recognition

Merits and Limitations

Page 29: Music recognition

Chart and data in Question 11,12,13 (for reference)

Yes37%

No63%

Are the users willing to help enlarge the database?

Page 30: Music recognition

OVER HALF (63%) of the respondents

refused to help enlarge the database

Yes37%

No63%

Are the users willing to help enlarge the database?

Page 31: Music recognition

Limitation

Over half of the respondents refused to help enlarge the database

As the enlargement of database relied heavily on users’ contribution

The database can not be enlarge

thus, searching result may not be desirable

Page 32: Music recognition

Why users love

15%

24%

28%

13%

9% 9%

3%

Reasons of attractions

Large database (a big song storage)

Very convenient

Free

Don't need to know much information about the song

Results are comprehensive.(with lyrics and youtube link)

Results are accurate

Other reason:

Page 33: Music recognition

Why users love

Reason of “Free” and “convenient” made up HALF (51%) followed by “Large database”. Two options regarding “Results” constituted only 9% each.

“Free” and “convenient” are the main reasons of attraction followed by “Large database”. On the other side, users don’t care much about the results.

The merits should be

Page 34: Music recognition

Why users refuse to use

18%

27%10%

6%

35%

4%

Reasons of why people refuse to use

Small databaseResults are not accurateAlways sing out of pitchNo smartphoneCharge for US$5Other reason:

Page 35: Music recognition

Why users refuse to useReason of “Charge for US$5” accounted for 35%

while “ inaccurate results” made up 27%, followed by “Small database”(18%). “No smartphone constituted the least ( 6%) while “always sing out of pitch” had 10%.

What the users considered the most is the CHARGE, followed by inaccurate results and small database. And users are slightly concerned about if they would sing out of pitch.

Page 36: Music recognition

Merits and limitations

The merits are mainly convenient and with large database. Yet, some respondents said that they wouldn’t help to enlarge the database. That’s the conflicts

And may have a vicious cycle: Database is not large enough and therefore the results will

not be very accurate, then less people to use, thus database cannot be enlarged

All users may leave because of the small database.

Besides, users always consider about money. If the app is charged for US$5, users may use GOOGLE instead. It’s because there share the similar function but music recognition is just more convenient.

Page 37: Music recognition

THANK YOU