A machine learning project does not have to be complicated or expensive. This is demonstrated by the solution we were able to provide to Drowgoo (http://drowgoo.nl). For a project to reduce pigeon damage without shooting, Drowgoo works with sensors that automatically recognise geese - and now also pigeons. This required an AI solution to recognise pigeons that would require as little computing power as possible.
Lightweight noise recognition modelling is precisely Aquila Ecology's expertise. Therefore, an AI model could be delivered with only a day's work. This model can distinguish 4 species of pigeons, or the absence of pigeon noise. Only the cave pigeon found the AI model slightly more difficult, probably because the sound of the cave pigeon (a rather low how-how) is more similar to most background sounds than the cooing of the other species.
Apart from the pigeon project, Drowgoo and Aquila Ecology are also collaborating in the field of bats. The combination of sensors developed by Drowgoo and Aquila Ecology's advanced AI model could make bat research much easier in the future.
