In the wild, there are less than 4,000 people of this species. The International Union for the Conservation of Nature (IUCN) has called the species “extremely endangered.”
The snow leopard is a huge cat found in the mountains of Central and South Asia. Due to their isolated location and elusiveness, these beautiful cats are difficult to examine. Snow Leopard Trust trap cameras took more than one million pictures to help save the species. This is how they did it:
I hear a chuff as a snow leopard approaches me. “How could you come so near to me without my knowledge?” he asked, his eyes broad and wide. They overestimate their invisibility. A ribbon-like turn and he was gone, “says Koustubh Sharma of the Snow Leopard Trust. But those trap cameras use heat and motion sensors. So they photograph more than just the elusive leopards. Imagine sorting through hundreds of thousands of photos by hand.
So here comes the AI. A new Microsoft AI solution is speeding up the research team’s procedure. According to Microsoft News, the system can sift through hundreds of thousands of photos in minutes. Sharma remarked, “You’re practically allowing snow leopards to snap selfies.” To save them, you need to know where they are and how many there are. But occasionally, you end up with thousands of irrelevant images. That’s where AI comes in. “
The developing developers hope the technology can soon identify a single person from images. “Is the snow leopard in Picture 1,240 the same as in Picture 1,000,240?” How do you connect these images? The current method is laborious and error-prone, and adding photos is like putting pieces into a 40,000-piece jigsaw puzzle. “We expect deep learning can help us locate matches,” said Mark Hamilton, the model’s creator.
Scientists need knowledge about the species to save the snow leopard. So they need lots of images. “More research,” Sharma says. “AI can help us estimate global snow leopard populations.” Watch the video below:
