Joe Cincotta

I am a computer scientist with expertise in machine learning, data science and human-centred design. After 15 years of working in software design for startup and corporate clients, I found my love for academia and scientific research after a stint as the technical lead in a medical imaging research group at the University of Sydney.

In 2020 I was fortunate to have the opportunity to work with the Welfare Conservation and Science team at Taronga, and we continue to collaborate on projects. This collaboration evolved into my current academic research at UNSW.

Current areas of focus:
- Applying machine learning research to wildlife conservation and animal welfare
- Predictive analytics in healthcare

Project: Autonomous identification and classification of animal behaviours in captivity using computer vision and machine learning methods

Supervised by: Shinichi Nakagawa, Arcot Sowyma, Richard Kingsford

Project Description: There are several research teams around the world currently evaluating AI to support the monitoring and assessment of animals in zoological settings. This data-driven understanding of animal behaviour can assist in optimising enrichment programs, habitat design and feeding strategies.

My research seeks to explore animal behaviour and welfare using AI and computer vision to identify and track individual animals throughout their daily lives, dynamically generating heatmaps, ethograms and behaviour budgets; essential tools in welfare assessment. Using an ensemble of machine learning methods and low-cost imaging sensors, this work seeks to use novel methods to collapse the latent space of imaging sensors to derive the essential information about the animals under observation, and also to build a model of those animals' behaviours, including individuation, using unsupervised models.