YomeciLand

BACKGROUND Recent advancements in the capabilities of computers to learn and adapt to data autonomously have provided new opportunities for creative practitioners using interactive and digital systems (Kogan 2016, Fiebrink 2019). While predominantly the domain of computer scientists and engineers, the potential of machine learning has recently received significant attention across art practice, games and digital media. With the prevalent use of sensors in interactive systems that can track data in physical space, models for how machine learning can recognise, analyse and classify this data in new and useful ways are emerging. However, without many existing works that demonstrate this application, its significance and potential is not well understood. CONTRIBUTION In response, YomeciLand investigated how machine learning could aid the capacity of audio recognition within a sound responsive interactive installation. Using the sound input of participants, the work evolves according to these playful and improvised sonic performances. YomeciLand demonstrates a model for the usefulness of machine learning as a new and creative computational technique for aiding audio recognition in sonic and interactive arts practice. SIGNIFICANCE Bunjil Place Gallery’s exhibition program aims to present and promote the best Australian and international art that engages, challenges and inspires audiences. YomeciLand was commissioned by Bunjil Place and the Casey City Council for a public exhibition from 26th September to 6th October. During this ten-day period nearly YomeciLand drew almost 5,000 visitors making it one of the most attended exhibitions in the Bunjil Place Studio. A major adaptation of YomeciLand was selected for Experimenta: Lifeforms, a national touring exhibition in 2020 – 2023.