GSoC 2025 Project #2 HarmonyHub :: Using Generative AI for adaptive learning in music (350h)

Mentors: Alberto Acquilino <alberto.acquilino@mail.mcgill.ca>, Mirko d’Andrea <mirko.dandrea@gmail.com>, and Keerthi Reddy Kambham <reddykkeerthi@gmail.com>

Skill level: Advanced

Required skills: Proficiency in Python and JavaScript (experience with Angular and Ionic is a plus); familiarity with generative AI frameworks (e.g., OpenAI GPT, Hugging Face); experience with web development and UI/UX design; knowledge of audio processing libraries, such as Essentia (optional); basic understanding of music theory and notation (optional).

Time Commitment: Full time (350 hours)

Forum for discussion

About: HarmonyHub is an ongoing open-source project that aims to bridge this gap by developing a modular, web-based platform for music education. The platform leverages modern web technologies (Ionic and Angular) to create reusable components and services, enabling a hybrid mobile app that can run in browsers or be deployed as mobile apps for iOS and Android. By integrating generative AI, HarmonyHub will empower music educators to create adaptive, personalized learning experiences for students, making music education more inclusive, diverse, and effective.

The HarmonyHub platform is built using the Ionic framework and Angular, enabling the development of a hybrid mobile app that can run directly in browsers or be deployed as native mobile apps on iOS and Android devices. Its modular architecture allows for reusable components and services, simplifying customization and scalability.

Generative AI will be used to:

  • Assist music teachers in defining exercises for their students.
  • Enable a code-free interface for creating and compiling custom web applications for music education

Aims: This project envisions integrating large language models (LLMs) and Generative

AI to empower music educators with tools for dynamic, adaptive learning. This will enhance student engagement, improve learning outcomes, and provide educators with powerful tools to tailor instruction to individual needs. The project will also contribute to the broader field of adaptive learning by demonstrating how generative AI can be applied to domains with complex, non-textual data like music.

Objectives:

  • Extend HarmonyHub to incorporate Generative AI for adaptive learning, focusing on creating personalized exercises tailored to students’ technical levels, age, and progression speed
  • Develop a user-friendly, code-free interface for teachers to customize and deploy exercises
  • Leverage Generative AI to analyze and enhance existing exercises, making recommendations for refinement based on student performance
  • Enhance inclusivity and diversity in music education by providing tailored learning experiences for students from varied backgrounds and abilities

What can I do before GSoC?

  • Familiarize yourself with the existing HarmonyHub codebase and its modular architecture
  • Explore generative AI frameworks (e.g., OpenAI GPT, Hugging Face) and their applications in adaptive learning
  • Experiment with audio processing libraries (e.g., Essentia, MIDI.js) to understand their potential for music education

Websites: Music Education Interface | Alberto Acquilino and GitHub - albertoacquilino/music-education-interface-ionic

Tech keywords: Generative AI, Adaptive Learning, Music Education, Angular, Ionic, Python, JavaScript, Audio Processing, MIDI

1 Like

Hey! am Jeeth Bhavesh Kataria from Ramaiah Institute of Technology .I’m eager to contribute to HarmonyHub: Using Generative AI for Adaptive Learning in Music as my Google Summer of Code project with INCF. As a computer science student with a strong interest in AI and neuroscience, this project excites me by merging generative AI with music education—a domain ripe for innovation. HarmonyHub’s potential to adaptively teach music aligns with INCF’s mission to advance computational tools, especially given music’s deep ties to brain function. My experience with Python, TensorFlow fuels my passion to develop an AI-driven tutor that personalizes learning while exploring its neuroscientific implications. I’m motivated to leverage tools like Magenta or transformers to create engaging, brain-inspired learning experiences. Through GSoC, I aim to enhance HarmonyHub, grow my skills, and contribute to INCF’s open-source community, making music education smarter and more accessible.

Hey, I’m Sampreeth, a pre-final year undergraduate student in Computer Science and Engineering. I came across the HarmonyHub project and found it incredibly exciting as it aligns perfectly with my research interests in generative AI and music generation.

I have hands-on experience in machine learning (ML) and deep learning (DL) and have published a paper in Springer on these topics. I have also collaborated with Intel on a project that leveraged advanced deep learning techniques. Currently, I am researching music generation and experimenting with different generative models such as VAEs and GANs to create adaptive and dynamic musical compositions.

From a technical standpoint, I am proficient in Python and JavaScript, with experience integrating APIs in real-world applications. I have also worked with Angular and Ionic, making me familiar with the tech stack used in HarmonyHub. Additionally, I have won three hackathons, where I successfully built and integrated APIs into functional projects, demonstrating my ability to develop scalable and innovative solutions.’

I am eager to contribute to this project by integrating generative AI into music education, enhancing adaptive learning experiences, and building a user-friendly, code-free interface for educators. I look forward to learning more about the project and collaborating with the team!

Hello everyone
I am really impressed by the HarmonyHub project—it’s such an inspiring merge of generative AI and music education! I am eager to contribute in the realms of Python/JavaScript development, exploring generative AI frameworks, and enhancing the UI/UX. I would love to work on the development of the platform’s adaptive learning features and support its mission of making music education inclusive and personalized. Also, I have integrated Google’s Flash 2.0 in my project Protochain to develop smart contracts on three different blockchains (Aptos, Polygon, Ethereum) based on user-specific prompts submitted by users.

I am eager to know how I can contribute?

hey @arnab1896 !
I wanted to let you know that I’ve successfully solved the multi‐color theme issue on github repo url you mentioned (GitHub - albertoacquilino/music-education-interface-ionic)
here’s the issue link : Improving UI/UX by adding the multi color theme · Issue #9 · albertoacquilino/music-education-interface-ionic · GitHub

although I only know about react and tailwind, IOnic and Angular was bit of a tough concept to digest.

What’s the next step to work in order to discuss this project with you.

Are we building the project on the same platform i.e. (https://mei-trumpet.web.app/)
or need to make another platform.
Right now students are interacting with Trumpets. In order to scale this application we need to provide more sound instruments such as : wind instruments ( Flute, Clarinet) and bowed string instruments (Violin, Viola).

Hmmmm…seems interesting !

Resource : https://www.incf.org/sig/harmonyhub-web-based-platform-learning-variable-pitch-musical-instruments

Hi,
Thanks for your interest! We are starting the selection process. I recommend interested contributors to work on the application proposal template: GSoC-Proposal-Template - Google Docs

There are also open issues on the repository which are helpful to start getting familiar with the codebase.

Once you have filled the application document, you can reach out to all the mentors via email for feedback.

Best regards,
Alberto

Hi Sampreeth,
Thanks for your interest! We are starting the selection process. I recommend interested contributors to work on the application proposal template: GSoC-Proposal-Template - Google Docs

There are also open issues on the repository which are helpful to start getting familiar with the codebase.

Once you have filled the application document, you can reach out to all the mentors via email for feedback.

Best regards,
Alberto

Hello everyone,
Thanks for your interest! We are starting the selection process. I recommend interested contributors to work on the application proposal template: GSoC-Proposal-Template - Google Docs

There are also open issues on the repository which are helpful to start getting familiar with the codebase. We will keep revising the pull requests on the go, according to our availabilities.

Once you have filled the application document, you can reach out to all the mentors via email for feedback.

Best regards,
Alberto