COMP4801 Final Year Project 2023-2024

About the project

From time to time, people may hear music anywhere like retail shops and restaurants. While they are exposed to the music, they may only remember parts of the rhythm, pitch, or lyrics for that music. As a result, it is hard for those people to find the name of it or know all the details of it. It could be annoying to them if they want to replay the music or use the music in other occasions.

Although there are many practical applications on music recognition, only a few of them try to merge the traditional searching methods (e.g., search by lyrics, album) with the audio recognition methods. Therefore, this project will explore these methods and suggest a solution to perform music recognition that take audio and traditional searching parameters as input. It is expected that the result of this project can help public and content creators to find their desired music and get the music information for different usages (e.g., detection for copyrighted music pieces).

Project Schedule

Sep 2023Research and Preparation of Detailed Project Plan
• Research on current applications for music
• Research on current approach in audio recognition and music genre classification
• Research on other relevant information for writing the detailed project plan
Update on Project Webpage
• Detailed Project Plan upload
• UI design
1 Oct 2023Deliverables of Phase 1
• Detailed project plan
• Project web page
Oct-Dec 2023Speech and Melody Recognition
• Retrieving music dataset for recognition
• Exploration on data type conversion
• Exploration on architectures using CNN
• Model building
• Analysis and optimization
Music genre classification
• Exploration on data type conversion
• Exploration on architectures using CNN
• Model building
Nov-Jan 2024Preparation of First Presentation and Interim Report
• Interim report writeup
• Preparation of ppt and relevant aids
8-12 Jan 2024First presentation
21 Jan 2024Deliverables of Phase 2
• Preliminary implementation
• Detailed interim report
Jan-Mar 2024Music genre classification (Cont.)
• Model building
• Analysis and optimization
Music Prediction
• Integration of the models and search engine
• explore the possibility of generating music piece based on the audio input
• Testing, analysis and optimization
• Application polishing
Mar-Apr 2024Preparation of Final Presentation and Final Report
• Final report writeup
• Preparation of ppt and relevant aids
15-19 Apr 2024Final presentation
23 Apr 2024Deliverables of Phase 3
• Finalized tested implementation
• Final report
26 Apr 2024Project exhibition

Project Supervisor

Prof. Xu Dong
Department of Computer Science

Student Information

Cheng Ho Kin (Beng CS)
UID: 3035779612