← All student projects · Living with intelligent instruments

Visualising how an intelligent instrument changes over time

Design visualisations that show musicians how their AI model changes as they retrain it.

Level: Summer research, Honours, Masters
Prerequisites: Web development in JavaScript/TypeScript (Svelte experience is a bonus) and some data visualisation experience. COMP1720/COMP6720 is a good background.

Our NIME 2026 study (five musicians, one session each) suggested that visualising what an MDRNN is doing in real time can help musicians trust and understand it. But those visualisations show a single moment. When a musician retrains their model every week, they need to see how it has changed: what new data went in, which gestures it has picked up, and how its responses differ from last week’s model.

In this project you’ll design and build visualisations that compare datasets and models across sessions, as part of impsy-web, IMPSY’s browser-based version.

• Summer or one semester: comparison views for dataset growth and model behaviour, built and tested on existing multi-session performance logs.
• Semester 1 Honours/Master (alongside the multi-week study): the same, plus formative feedback sessions with musicians from the study during their final interviews.

Tags: #visualisation #XAI #interactive-machine-learning #web #IMPSY

How to apply

Contact Charles Martin with your CV, your unofficial transcript (if you are an ANU student), and a brief statement (200 words) explaining how you would approach this project. Before applying, read the Join page and our project expectations.