Over the course of three years, scientists set up 225 hidden cameras in the Serengeti to see what they could find without human interference. The result is 1.2 million images of more than 40 different types of wildlife out in the savannah.
In 2010, University of Oxford post-doctoral researcher Alexandra Swanson and her team setup cameras on trees and steel rods within a 400-square-mile area, replacing 15 percent of the cameras every year due to weather, insects, and even four-legged vandals, such as hungry hyenas or curious elephants. The cameras used heat and motion sensors to capture photographs of animals when they approached, producing up close and personal images of some of the most endangered and rare African animals in the wild, such as the Sumatran rhino, cheetah, and grey-faced sengi (an elephant shrew found only in Tanzania).
Swanson, her Harvard ecology post-doc partner Magaret Kosmala, and a group of 28,000 volunteers from over 70 countries refined the 1.2 million pictures with the citizen science web portal Zooniverse, and distilled the results into a collection of 300,000. That collection is now open to public access on an online archive called Snapshot Serengeti. Users can access any of the photos and search by image tags to find specific animals or actions they want to see. The "selfie" tag features animals, such as a baboon or Grant's gazelle, looking into the camera, "young"-tagged images produce a series of baby animals, and "kill" shows the Serengeti's predators feasting on less fortunate speccies. There are also instant animal gifs, thanks to the cameras taking three-shots bursts of moving animals, such as a herd of elephants.
Snapshot Serengeti has already classified its project data, but there are still active images left to identify in the archive if you want to help. You don't need to know a diki dik from a duiker, the classify tool takes you through a series of filters for the characteristics of each animal, such as, "Are its horns curved or straight?" and, "Does it have spots or does it have stripes?" With the filters, you identify some of the rarest animals on the planet. Kosmala and Swanson tested the accuracy of their method by giving 4,000 of the online images to five experts. Citizen scientists classifying online matched the experts nearly 97 percent of the time.