Sohyun Im

Research Assistant

Queen Mary University of London

About Me

Sohyun Im is a Research Assistant at the Centre for Digital Music, Queen Mary University of London, where she is working on the Universal Acoustic Vision project in partnership with Meta Reality Labs. Her current research focuses on multichannel audio processing, acoustic imaging, machine learning, and real-time spatial audio systems for wearable devices.

She received her Master of Science in Sound and Music Computing from Queen Mary University of London, where her dissertation explored music structure analysis using machine learning. She also holds a Bachelor of Science in Sound Engineering from the University of West London, where her final-year project focused on virtual analogue modelling.

Outside her research, Sohyun has been actively involved in two London-based communities: Dynamic Cast, a peer-to-peer C++ study group supporting people underrepresented in technology; and the Oxford and Cambridge Musical Club, where she performs as a classical pianist.

Sessions

  • Build Your Own Shazam

    Audio Fingerprinting from First Principles
    10:00 - 13:00 UTC | Monday 9th November 2026 | Empire
    Beginner
    Intermediate
    In-Person Only

    Audio fingerprinting is one of the most successful real-world applications of audio signal processing. It powers music recognition systems such as Shazam, enabling reliable identification of songs from short, noisy recordings in a few seconds. While many audio developers are familiar with audio fingerprinting in practice, the underlying techniques are often encountered only at a high level. In this hands-on workshop, participants will build a complete audio fingerprinting engine based on Avery Wang’s landmark 2003 Shazam paper. Starting with raw audio, we will progressively construct the identification pipeline: generating spectrograms, detecting robust spectral landmarks, creating constellation maps, producing landmark hashes, […]

  • Workshop: Practical Machine Learning

    Embed a generative AI model in your app and train your own interactions with it
    09:30 - 12:30 UTC | Monday 11th November 2024 | Bristol 3
    Beginner

    In this workshop we’ll explore the fundamentals of Machine Learning. We will run through an easy to follow machine learning model that will: Be easy for beginners Run on the CPU Be real time This will cover an intro to Machine Learning, small vs large models and an introduction to a training environment in python. We aim to make this workshop as interactive as possible, with the idea of having a trained model in session for everyone to use/play with. This will be a self-contained workshop aiming to be accessible to all levels of learning - all elements used in […]

  • An Introductory Guide to Virtual Analog Modelling

    The Intersection of Analog and Digital Audio Processing
    11:20 - 12:10 UTC | Wednesday 13th November 2024 | Empire
    Beginner

    Audio circuits, such as guitar pedals and amplifiers, process input signals to output signals with specific audio effects. The process of replicating the electrical behaviour of these circuits in a digital environment is known as Virtual Analog (VA) modelling. This session aims to provide a basic understanding of VA modelling and explore various methods to achieve it. Finally, using these methods as a foundation, we will look at modelling the circuit of the MXR Distortion+ guitar pedal that emulates the circuit’s behaviour.