| Summary: | HABIT (Horse Automated Behaviour Identification Tool) is an Animal Computer Interaction (ACI) project, on the interdisciplinary boundary between equitation science and computer science. HABIT will automate the analysis and recognition of horse-to-horse and horse-to-human behaviours, as observed in unconstrained / ad-hoc video. A horse-to-horse dyad video dataset will be compiled, illustrating exemplar behaviours. Behavioural signatures will be manually identified from video. Next, a system will be developed and trained to recognise these signatures. The tool will then be evaluated, when applied to both horse-to-horse and horse-to-human video clips. In the study of animal behaviour, an ‘ethogram’ is a set of comprehensive descriptions of the characteristic behaviour patterns of a species. HABIT is potentially the first step towards the ‘automated ethogram’. This project provides a welfare-orientated approach to evaluating horse behaviours. When horses are handled, trained or ridden, HABIT will help ensure that these experiences occur within the natural repertoire of equine behaviours. There is also scope to engage and educate the public about horse behaviours; both for general interest and to raise welfare-awareness. Additionally, automation could play an important methodological role in animal-centred design by reducing human biases during the requirements and evaluation processes.
|
|---|