- This event has passed.
Computer Science: Understanding & Creating Sound using Large Language-Audio Models with Professor Wenwu Wang from University of Surrey
June 18 @ 17:00 - 17:45
Imagine if computers could listen to the world the way we do — hearing music, understanding speech, recognizing rain, or even telling a story from sound effects alone!
Today, scientists are teaching powerful AI systems called Large Language Models (LLMs) to understand and create sound. Just like AI can read and write text, it can now learn to “read” and “write” audio — from voices and songs to everyday noises like footsteps or thunder.
In this exciting talk, we’ll explore how AI combines language skills with sound skills to do amazing things, such as:
- Creating brand-new music and sound effects
- Describing what’s happening in an audio clip
- Separating different sounds (like pulling vocals out of a song)
- Compressing audio to store it more efficiently
- Turning sounds into creative stories
We’ll showcase cool projects like AudioLDM, AudioLDM2, and WavJourney, which can generate sounds and audio stories; AudioSep, which can separate mixed sounds; ACTUAL, which writes captions for audio; SemantiCodec, which smartly compresses sound; WavCraft, which helps edit and create audio content; and APT-LLMs, which help AI “reason” about sounds. We’ll also introduce special training collections like WavCaps, Sound-VECaps, and AudioSetCaps that help teach AI how to listen better.
Get ready to discover how AI is learning to turn sound into stories — and how the future might include computers that truly understand the world of audio!
Lecturer Profile: Wenwu Wang is a Professor in Signal Processing and Machine Learning and an Associate Head in External Engagement, School of Computer Science and Electronic Engineering, University of Surrey, UK. He is also an AI Fellow at the Surrey Institute for People Centred Artificial Intelligence.
Our events are designed so that a group of students can take part from a classroom or students and teachers can take part as individuals, either from school or from home. There is no charge to take part.
As a live, participative event, this session will not be recorded but slides and other resources will be shared after the event, where available.
Suitable for: Aged 16+ pre-university students studying, or with a possible interest in studying, Computer Science and related disciplines. Student groups as well as students and teachers joining individually are very welcome.
Preparation/Pre-Reading: Students may wish to consider the topic in advance and questions they have. Students can also think about questions about studying at university level.
The IT Setup: The event will be run on Zoom and participants will be able to join using laptops, tablets, and mobile devices. If you are joining as a school group, we recommend connecting with one device and projecting on to a whiteboard. You will be able to interact with the speaker via typed comments and questions and/or via a microphone. Full details will be sent to those who have registered (see below) prior to the event.
Booking a Place: This is free for schools and colleges. If you would like to take part in this event, please register your interest using the link to the registration form below: