In the movie “Rise of the Planet of the Apes”, the orangutans awakened by genetic drugs unite against humans and almost push humans into desperation. Such a story will only be staged in the movie, and the reality is just the opposite.
Because chimpanzees are the favorite pets of many rich people, they have always been favored by poachers. Since only young chimpanzees are suitable for pets, poachers usually kill adult chimpanzees in their families. The annual illegal trade amount of chimpanzees is tens of billions.
On social sites like Instagram, there are a lot of posts on chimpanzees. To save this endangered species, a new chimpanzee facial recognition system has been developed to help conservation organizations and law enforcement agencies detect chimpanzees that are illegally traded online.
The facial recognition system, called ChimpFace, was developed by the project’s founder, Alexandra Russo, and computer vision expert Colin McCormick. They collected thousands of chimpanzee photos from animal protection organizations for algorithmic training.
According to Alexandra Russo, the next step is to build an algorithm that identifies a specific chimpanzee and learn from the face recognition algorithms of companies such as Google and Facebook, because different chimpanzees face are completely different. This is bay far ChimpFace’s biggest challenge.
This technology was selected last year by Conservation X Labs’ Technology Awards, a company dedicated to user technology innovation to protect endangered species. Today, ChimpFace works with law enforcement and social media to develop new versions to get this technology into action faster.
Although face recognition technology has become popular in smartphones, and chimpanzees and humans are also primates, recognition is much more difficult than human faces.
Previously, Michigan State University had developed a facial recognition system PrimNet for primates. Researchers have used face recognition technology on these primates, and there has been an “overfitting” problem. That is to say, the details are not obvious enough.
This is because the animal’s hair and eye color characteristics are quite different, so the research team had to manually mark the animal’s eyes and mouth to assist the system identification.
For artificial intelligence-based ChimpFace, it is also inseparable from humans. Computer vision expert Colin McCormick said that after the algorithm marks possible posts and photos on the Internet, it still needs manual review.
Although the facial recognition of the targeted animals is not mature enough, the application of similar technologies has been increasing. For example, PrimNet, just mentioned, can track these animals with Android apps to help forest conservation personnel better protect some endangered primates with an accuracy of over 90%.
In the field of farming, sheep face recognition and face recognition have also begun to be applied. In China, Alibaba and JD have introduced pig face recognition technology last year. This technology can be used to observe and record the weight, growth and health of each pig.
According to the United Nations Environment Program, about 3,000 people are illegally trafficked every year. Since 2005, some 30,000 people have died in poaching, hoping that technology can change their destiny.