Supervised Learning Computer Vision Benchmark for Snake Species Identification From Photographs: Implications for Herpetology and Global Health

Article Article scientifique Article Supervised Learning Computer Vision Benchmark for Snake Species Identification From Photographs: Implications for Herpetology and Global Health eng 2021 author Durso Andréw Michaël 909495 author Moorthy Gokula Krishnan author Mohanty Sharada P. author Bolon Isabelle 0000-0001-5940-2731 114399 author Salathe Marcel 845432 author Ruiz De Castaneda Rafael Luis 789107 10.3389/frai.2021.582110 33959704 2624-8212 https://www.frontiersin.org/articles/10.3389/frai.2021.582110/full We trained a computer vision algorithm to identify 45 species of snakes from photos and compared its performance to that of humans. Both human and algorithm performance is substantially better than randomly guessing (null probability of guessing correctly given 45 classes = 2.2%). Some species (e.g., Boa constrictor) are routinely identified with ease by both algorithm and humans, whereas other groups of species (e.g., uniform green snakes, blotched brown snakes) are routinely confused. A species complex with largely molecular species delimitation (North American ratsnakes) was the most challenging for computer vision. Humans had an edge at identifying images of poor quality or with visual artifacts. With future improvement, computer vision could play a larger role in snakebite epidemiology, particularly when combined with information about geographic location and input from human experts. Fine-grained image classification Crowd-sourcing Reptiles Epidemiology Biodiversity Dewey 613 Autre Fondation privée des Hôpitaux Universitaires de Genève (award QS04-20) 66 Institut de santé globale 22A1 false Frontiers in Artificial Intelligence 4 b1bdec17-a36a-4656-bba2-d379eb798322 Article (Published version) Public Dursoetal2021SupervisedLearningfrai-04-582110.pdf 3142797 application/pdf CC BY 5ba1179d-032a-4df5-83b4-b24d526758d9 Appendix Public PMC8093445_SupplementaryFiles.zip 966458 application/zip openAccess Faculté de médecine / Section de médecine clinique / Département de santé et médecine communautaires / Institut de santé globale

Title	Published in	Access level	OA Policy	Year	Views	Downloads

Supervised Learning Computer Vision Benchmark for Snake Species Identification From Photographs: Implications for Herpetology and Global Health	Frontiers in Artificial Intelligence			2021	242	364

Archive ouverte UNIGE

Salathe, Marcel