Future of Music And AI
Caroline Pegram / Beth Shulman from Audi Intelligence Agency (AIA) will talk about how emerging technologies like AI & blockchain are upending existing business models. This is causing concerns with both artists and record labels that tech will bring the industry to its knees all over again. How can frontier technologies be used in a way that expands creative and financial opportunities for artists & industry, instead of bypassing them?
Learning more with less
Flora Salim will talk about how we can improve how we train ML models with smaller training data, fewer samples, and less resources in the dynamic environments.
The proliferation of sensors and Internet of Things leads to new opportunities and challenges for human behaviour modelling and forecasting behavioural patterns at scale. Most representation learning techniques require a large amount of well-labelled training sets to achieve high performance. The high expense of annotating behaviours of interest and the heterogeneity of both the data sources and the downstream tasks have led to the rising popularity of data-efficient learning approaches that require no labelled data, minimal examples, and are generalisable. Our recent data-efficient learning techniques to achieve efficient and compact representations from heterogeneous sensor data will be briefly covered. Several open issues including explainability and robustness in dynamic environments will be discussed.
For over two decades, Caroline Pegram has worked as a producer for award winning companies and celebrated media personalities, with a specialisation in science communications. Most notable is her 20 years as producer for Dr Karl Kruszelnicki.
Always leading with curiosity, she ventured into the world of Artificial Intelligence in 2018, working in Innovation & Strategy Director roles, with local music companies to help expand their technology businesses forming both commercial and academic global partnerships. In 2020 she led the team that won the very first global AI Song Contest. She recently joined the Audio Intelligence Agency as Director of Strategy. Caroline holds a Visiting Fellowship at UNSW and was a finalist in the Women in AIA Awards in 2022.
Originally based in NYC, Beth is a seasoned executive producer with a 20+ year background in film & advertising, with clients like Apple, Facebook & Beats by Dre. In 2019 She delved further into tech and co-founded Heyday, a lab that works in experimental prototyping and new ways of thinking with companies like Google. She is a co-founder of Audio Intelligence Agency, a product studio that brings artists & technologists together to build with emerging technologies and develop the future of music.
Professor Flora Salim is the inaugural CISCO Chair of Digital Transport, UNSW Sydney. Her research sits in the cross-cutting areas of ubiquitous computing, machine learning, and data science, with specific interests on representation learning of spatio-temporal and mobility behaviours and data-efficient learning with multimodal sensor data. Her recent research focus includes self-supervised learning, machine learning for multimodal time-series, explainable AI, fair machine learning, with specific applications on mobility data science and personalised recommender systems for smart cities/buildings/transport/energy, smart environments, and intelligent task assistants. Her research has been funded by Australian Research Council (ARC), Victorian Government, Microsoft Research US, Northrop Grumman Corporation US, Rheinmetall Defence Australia, Qatar National Priorities Research Program, IBM Research, Alexander von Humboldt Foundation, Bayer Foundation, several city councils, and many other industry and government partners/funders. She is the recipient of Women in AI Awards 2022 Australia New Zealand in the Defence and Intelligence category. She has received several fellowships in the past, including Humboldt Fellowship, Humboldt-Bayer Fellowship, Veski Fellowship, and ARC Postdoctoral Industry (APDI) Fellowship. She obtained her PhD from Monash University in 2009.
