Can AI text signals help save snow leopards from extinction?

Snow leopards cannot growl. So, when we step on one of these fierce predators, she is gloating.

“Wonderful,” as she calls her, she was orphaned 12 years ago in Pakistan Gilgit-Baltistan administered by Pakistan.

After years of reading the staff to feed her, she does not know how to hunt in the wild – and cannot be released.

“If we release her, she would just attack a farmer sheep and be killed,” the watching of Lovely, Tehzeeb Hussain, tells us.

Despite the laws that protect them, between 221 and 450 snow leopards are killed every year, says the World Nature Fund (WWF), which has contributed to a 20% decrease in the world population over the last two decades.

More than half of these deaths were revenge for the loss of livestock.

Scientists now believe that only 4000 to 6000 snow leopards are left in nature – with approximately 300 of them in Pakistan, the third largest population in the world.

To try to reverse these anxious trends, WWF – with the help of Pakistani University at Lahor University of Management (Lums) – has developed cameras powered by artificial intelligence (AI).

Their goal is to find the presence of a snow leopard and warn the peasants via a text message to move their livestock safe.

Tall, with a sun panel mounted on top, the cameras are located high in the fruitless and rough mountains of nearly 3,000 m (9 843 feet).

“The territory of the snow leopard,” says Asif Ikbal, a conservationist from WWF Pakistan. He walks us with a few more steps and points to watch on the ground: “They are quite new.”

Asif hopes that this means that the camera has recorded more evidence that the AI ​​software – which allows it to distinguish between humans, other animals and snow leopards – works.

Currently, WWF is testing 10 cameras located in three villages in Gilgit-Baltistan. It took three years to train the AI ​​model to find these categories with impressive – if not perfect – accuracy.

After we get back down the mountain, Asif pulls off her computer and shows me a control panel. I'm there, in a series of GIF. He correctly discovers that I am human. But as we scroll down into the list, I appear again and this time I am referred to as a human and animal. I wear a tight white fleece, so I forgive the program.

Then, Asif shows me the shot of money. This is a snow leopard recorded several nights before, in a night vision. He picks up another week before. This is a snow leopard raising its tail against the nearby rock. “This is Mother Leopard, it looks like she marks her territory,” says Asif.

Adjusting the cameras in rocky, at a height of height, it took many attempts and mistakes. WWF has gone through several types of batteries until it finds one that can withstand the harsh winters. Specific paint has been chosen to avoid reflecting light as the animals pass.

If the cell service fails in the mountain, the device continues to record and capture data locally. But the team had to accept that there were some problems they just couldn't solve.

While the camera lens was protected from a metal box, they had to replace the solar panels damaged by landslides.

Not only has technology caused problems. Obtaining buying the local community was also a challenge. At first, some were suspicious and doubted whether the project could help them or snow leopards.

“We noticed that some of the wires were cut off,” Asif says. “People had thrown blankets over the cameras.”

The team also had to take into account the local culture and the emphasis on the personal life of women. The cameras had to be moved because the women were walking too often.

Some villages have not yet signed forms of consent and confidentiality, which means that the technology cannot yet be introduced in their field. WWF wants a binding promise that local farmers will not access poachers to the footage.

Sitara lost all six of its sheep in January. She says she has taken them to graze over their home, but that a snow leopard has attacked them.

“It was three to four years of hard work that raised these animals and everything ended in a day,” she says.

The loss of the livelihood left her in bed for a few days. Asked if she hopes that AI cameras can help in the future, she replies: “My phone is almost not a service during the day, how can text help?”

At the collection of the elders of the village, the leaders of the village of Hiber explain how the attitudes have changed over the years and that the growing share of their village understands the importance of snow leopards and their impact on the ecosystem.

According to WWF, snow leopards hunt Ibex and blue sheep, which stops these animals from excessive and helping to keep lawns so that the villagers can feed their livestock.

But not everyone is convinced. A local farmer questions the benefits of animals.

“We had 40 to 50 sheep before, now we have only four or five. The reason is the threat of snow leopards and Ibex to eat the grass,” he says.

Climatic changes also play a role to play why some feel threatened by snow leopards. Scientists say warming temperatures have led the peasants to move their cultures and livestock to higher areas in the mountains, reaches to their own habitat of snow leopards, making cattle more than a goal.

Whether the peasants are convinced of the conservation message or not, WWF tells us that legal penalties have served as a strong deterrent in recent years. Three men were imprisoned in 2020 after killing a snow leopard in the Hopper Valley, about two hours by a car from Hiber. One of them had posted pictures of himself with the dead animal on social media.

Although those involved in the camera project hope that their AI devices can influence, they know that they cannot be the only solution.

In September, they will start to try out the smells, sounds and lights in the camera sites to try to prevent the snow leopards from moving to nearby villages, and to put livestock in danger.

Their work, which traces these “ghosts of the mountains”, is not over yet.