DeepMind has joined the likes of those who are using AI to save the fragile wildlife population on a grand scale. The company is collaborating with conservationists and ecologists on a project to save wildlife by using an AI that speedily detects and counts animals in "millions" of photos taken over the past nine years in Tanzania's Serengeti National Park, revealed a report in Endgadget. Notably, unlike the prolonged time required for volunteers to return labelled photos, DeepMind has developed a model than can label most animals almost as well as humans, shortening the process by up to nine months.
The development becomes more important considering animals seldom cooperate with motion-sensitive cameras. The AI on its part can now recognise out-of-focus felines or fast-moving ostriches to label the animals. The Endgadget report further elaborates that DeepMind is developing a pre-trained version of its AI Model that would only need "modest" hardware and little internet connectivity to do its work. Powerful computers and fast internet access usually turn out to not only disrupt wildlife but prove expensive to deploy as well. The DeepMind time hopes to validate its models soon by rolling them out in the field to find out how their product performs in real-life conditions.
The new development could significantly improve conservation efforts if it proves to be successful. With wildlife species becoming vulnerable in the Serengeti due to growing human encroachment, the machine system could not only track population distributions, movement and behaviour in detail but could quickly provide a date to conservationists so that they can respond in time to short term changes.
Comments
0 comment