Golden snub-nosed monkeys in China's Qinling Mountains on October 17, 2024 Photo: VCG

Though just three years old, monkey Fangtang (lit: sugar cube) is already the most "emotionally intelligent" and skilled food-beggar among a troop of 200 golden snub-nosed monkeys. His father, "Yangcong" (lit: onion), despite being 14 years old, in contrast, is a shy and calm gentleman who prefers lounging in the shade and watches his sons Fangtang and Kafei (lit: coffee) playing hide and seek.  

These storybook-like details in fact reflect the daily behaviors of a troop of golden snub-nosed monkeys in China's Qinling Mountains. The golden snub-nosed monkey is an endangered primate species endemic to China, and a flagship primate species in the Qinling Mountains. If not for an AI-powered animal recognition system that was recently created by a Chinese university team, it would have been nearly impossible to document the troop's vivid stories, let alone witness how technology is bringing smarter solutions to China's biodiversity conservation. 

Researcher Guo Songtao Photo: Courtesy of Guo Songtao

Digital ID for monkeys 

The animal recognizing system was developed by researchers at Northwest University, in Northwest China's Shaanxi Province. While lead developer Guo Songtao formally described the system as an "AI framework for automated quantitative measurement of animal behaviors," he revealed to the Global Times that it could essentially function as a "high-tech translator" of animal body language, and has solved a "long-standing challenge" in field observation. 

In the past, scientists had to monitor outdoor animals with their own eyes, a method limited by efficiency and accuracy. Even after video-assisted tools were introduced, storing all that footage and following fast-moving creatures remained difficult. 

With methods such as training AI and high-precision digital recognition, the system established by Guo's team can store massive amounts of real-time monitoring data and track the movements of wild golden snub-nosed monkeys. Taking Fangtang as an example, his food-begging behavior was found through tracking 17 digital spots projected on his body over a long, uninterrupted period.  

Including Fangtang, Guo's team has analyzed the behavioral patterns of around 200 golden snub-nosed monkeys and established "digital ID profiles" for each individual. On the profile, each monkey bears a unique name reflecting his or her behavioral traits. Take the individual dubbed "Scar" as an example. He plays the role of a leader in the group, his name a tribute to his combat prowess and fierce temperament. However, prolonged observation revealed that as he aged, his demeanor gradually mellowed.

"Studying animals shouldn't just be about knowing what they are, or how they look. We must deeply understand their behavior to develop better conservation strategies," Guo told the Global Times. 

The Chinese team's achievement has set a global precedent in primate research. Facial recognition for the species is nothing new, but the Chinese team's behavioral recognition system is "the first of its kind worldwide," said Guo, adding that the current outcome had undergone several years of polishing. 

Groundbreaking research

Back in early 2018, Guo's team had started fusing AI technologies into primate research. In 2020, Tri-AI, a facial recognition system made its debut, heralding the world's first research tool of its kind. The development of this project was "closely linked to China's technological advancement and independent innovations," said Guo. 

"Whether in automotive manufacturing or other fields, the rapid progress in [China's] AI technology has allowed us to identify synergies with our specialized expertise," the researcher noted. 

Despite achieving two world-first milestones, the team continues to deepen the fusion of technology and wildlife conservation, and is particularly working hard to make scholarly tools available to teams and organizations worldwide. 

Currently, the research team has made its latest system an open-source tool - including both the framework and its coding - sharing it with the broader academic community. A team from Indiana University in the US has obtained authorization and is currently adapting this technology to study ring-tailed lemurs.

Including the ring-tailed lemurs, the system has now been applied across 10 different species. Beyond primates, its applications extend to hippopotamuses, alpacas, red pandas, and more. 

Noting that wildlife conservation should not be exclusive to academic circles, Guo and his team have developed a WeChat mini-program called "Monkey Face Recognition." By registering, the public can not only access the "digital ID profiles" of existing golden snub-nosed monkeys, but also upload wildlife footage to contribute to species identification and research.

The 30th International Primatological Society Congress (IPS 2025) was recently held in Antananarivo, Madagascar. As one of 46 Chinese scholars in attendance, Guo presented his team's two keynote reports, including the golden snub-nosed monkey communication behavior recognition technology to international scholars. 

He noted at the event that the international community can grasp "China's evolving wisdom and resolve in species conservation" through sharing projects. Also, he revealed to the Global Times that his team is currently developing another "primate research breakthrough" that "will stun the international academic community." 

In 2026, the Chinese city of Xi'an will host IPS 2026. Guo said he hopes to unveil his new research at the conference. "With fast-developed technologies, we hope that we can drive the world's new wave of progress in animal behavior research and accelerate the pace of innovation," Guo told the Global Times.