There has been much public discussion about the increased use of AI, much of it fairly negative focusing on potential threats to employment &/or to society itself (if AI develops sufficiently to think for itself & so effectively does away with the need for humans).

However, there are valid & worthwhile uses for AI & using it to analyse large quantities of data quickly & accurately is a tangible benefit.  AI has been developed which can identify many categories of marine litter successfully, based on analysis of drone imagery from mainland beaches.  Unfortunately, as the composition of island marine litter differs significantly from shore-based beaches, there is a knowledge ‘gap’ & consequently, the algorithms applicability for island beach litter analysis is limited.

This project seeks to fill that gap using drone imagery from island beaches cross-referenced with items manually collected from beach cleans to train the AI algorithms to recognise & identify the fishing & aquaculture items most commonly found on island beaches.

 

This grant is towards the costs for equipment & staff time.