BAMBI - Detection Dataset

This dataset contains thermal drone imagery for the purpose of object detection of wildlife, specifically focusing on red deer (Cervus elaphus), roe deer (Capreolus capreolus), and wild boar (Sus scrofa). The data was collected using aerial thermal sensors under field conditions and is intended to support research in automated wildlife monitoring, species differentiation, and animal behavior analysis. The dataset comprises a total of 19252 video frames mostly showing wildlife, divided into a train set with 15,730, a validation set with 1,696 and a test set with 1,826 images. The images are split based on individual videos and flight locations to avoid biases related to environmental or flight characteristics. As in the given use case, we are not focusing on the classification of the species; they are summarized as one "animal" label. The dataset is also available on Roboflow: https://app.roboflow.com/dseducation/bambi-bounding_box-20250523/1