APA Detecting Rats with a modified yolov8 net
The Active Place Avoidance rig needs to automatically detect if a rat enters a predetermined region of the arena.
This video shows an early attempt to detect rat location in real time. It uses opencv with a modified yolov8 neural net, trained on a SIDB dataset of approx 537 labelled images (originally around 173 images, which were modified to increase the number of images)
The dataset is available at https://app.roboflow.com/sidb-worksho...
For more information on our system of modular Raspberry Pi rigs for behavioural neuroscience, visit https://github.com/Kind-Wyllie-lab/sa...