Aerial Elephant Dataset (AED) and Semmacape Dataset (SD)

doi:doi:10.57702/fan5wzxq

Aerial Elephant Dataset (AED) and Semmacape Dataset (SD)

Two aerial image datasets are studied in this work. The first one is the Aerial Elephant Dataset (AED) which was proposed by [6] to promote research on animal detection in practical conditions. This dataset consists of 2101 aerial images with dot annotations for a total of 15,511 African bush elephants in their natural habitat, with complex backgrounds, as seen in Fig. 1. The second one is the Semmacape Dataset (SD) [7, 8] which contains 165 box-annotated aerial images collected in the Gironde estuary and Pertuis sea Marine Nature Park, France, during spring 2020.

BibTex:

@dataset{Tanya_Singh_and_Hugo_Gangloff_and_Minh-Tan_Pham_2025,
    abstract = {Two aerial image datasets are studied in this work. The first one is the Aerial Elephant Dataset (AED) which was proposed by [6] to promote research on animal detection in practical conditions. This dataset consists of 2101 aerial images with dot annotations for a total of 15,511 African bush elephants in their natural habitat, with complex backgrounds, as seen in Fig. 1. The second one is the Semmacape Dataset (SD) [7, 8] which contains 165 box-annotated aerial images collected in the Gironde estuary and Pertuis sea Marine Nature Park, France, during spring 2020.},
    author = {Tanya Singh and Hugo Gangloff and Minh-Tan Pham},
    doi = {10.57702/fan5wzxq},
    institution = {No Organization},
    keyword = {'Aerial Images', 'Remote Sensing', 'Wildlife Counting'},
    month = {jan},
    publisher = {TIB},
    title = {Aerial Elephant Dataset (AED) and Semmacape Dataset (SD)},
    url = {https://service.tib.eu/ldmservice/dataset/aerial-elephant-dataset--aed--and-semmacape-dataset--sd-},
    year = {2025}
}