Baltic Seabird Dataset
Annotations: The annotations provide the bounding boxes of the birds present on the ledge classifying them into Adult birds, Chicks and Eggs.
Quality of the dataset: real-world, challenging, class-imbalanced dataset.
Use cases: An AI which can detect/predict behaviors like copulation, parenting, fights etc. can help researchers understand and automate the behavior analysis process. An “Event Flagging” system which can predict and flag the behaviors like arrival and departure of the guillemots from the ledge can be useful to understand the Baltic Sea nutritional statistics.