To widen the impact of the Cat Royale project, members of the Somabotics Fellowship are using expert knowledge of Computer Vision and Animal behaviour to produce a new dataset to enhance understanding of Cat behaviour and improve animal welfare.

The team has leveraged cutting edge AI to distill hundreds of hours of raw video footage into a format for users across disciplines to use in feline research. The Cat Royale Comprehensive Dataset (CRCD) will help create the next generation of animal understanding AI, training future AI models to recognise the body-language, mood and behaviours of cats. Thanks to these advances artificial intelligence it will be used increasingly to monitor animal health and wellbeing.

For this project, we have been joined by Joseph Ho, an animal behaviourist specialising in the behaviour of cats and dogs who will be providing input on animal behavioural problems. Within Cat Royale, Joseph applies his expertise to the detailed annotation of feline behaviour, helping to capture the subtle nuances of body language and emotional expression. His input ensures the dataset not only supports cutting-edge AI research but also reflects a deep commitment to improving cat welfare.

The dataset, and an associated publication will be published later this year, and will allow the legacy of Cat Royale to extend far beyond the duration of the project. Along with the dataset itself, the software tools designed for and used in the creation of CRCD will be released, allowing future development of the dataset.

Ahead of its release, Somabotics will hold a seminar bringing together potential users and stakeholders in the dataset from fields of Animal Sciences, Computer Vision, Robotics, HCI and Art to share information on its upcoming release and connect CRCD with its cross-disciplinary audience.