The iWildCam 2020 Competition Dataset

• DOI: 10.48550/arxiv.2004.10340
• Publication Date: 2020-01-01
• Authors: Sara Beery; Elijah Cole; Arvi Gjoka
• Keywords: Leverage (statistics); Camera trap; Citizen Science; Data set; Scope (computer science)

Camera traps enable the automatic collection of large quantities image data. Biologists all over world use camera to monitor animal populations. We have recently been making strides towards species classification in trap images. However, as we try expand geographic scope these models are faced with an interesting question: how do train that perform well on new (unseen during training) locations? Can leverage data from other modalities, such citizen science and remote sensing data? In order tackle this problem, prepared a challenge where training test different cameras spread across globe. For each camera, provide series imagery is tied location camera. also set seen our The correctly classify traps.