Early prediction of bumblebee flight task using machine learning

A comparison of machine-learning assisted optical and thermal camera systems for beehive activity counting

“…These are cases when the information from either test camera did not provide a clear flight category determinant. Similar to our previous work, machine learning is based upon the motion characteristics of the bee, which is used to classify each flight [20] . The use of a camera rather than radio-frequency technology, as in the previous study, allowed for shape metrics to be extracted, increasing the depth of machine learning used.…”