Abstract

With the advent of modern transient surveys and the upcoming Rubin Observatory soon to start full operations, time-domain astronomy has firmly entered its era of big data, and has moved into a domain where we are flooded with novel transient discoveries rather than artifacts, thanks to machine learning methods. Prioritisation of finite follow-up resources and early classification of objects are crucial to make the most of this.

In this talk I will discuss the ongoing development of high-performance deep learning classifiers and contextual classification algorithms for the Gravitational-wave Optical Transient Observer (GOTO)—an all-sky optical transient survey built specifically for real-time follow-up of gravitational wave (GW) events. I will also discuss Kilonova Seekers, a citizen science project hosted on Zooniverse which shows data from GOTO to members of the public in real time – uniquely enabling them to join the search for GW counterparts, and make impactful scientific discoveries in the process, whilst also delivering vital data for real-time adaptive classifiers and active learning.