Advertisement
China technology
TechPolicy

Inside China Tech: tracing the historical roots of recent AI, moneylending innovations

  • A video game developer has used artificial intelligence to recreate the faces of ancient Chinese emperors
  • China’s online microlenders put a digital twist on the age-old practice of moneylending

Reading Time:4 minutes
Why you can trust SCMP
A painting of Jiaqing Emperor, the sixth emperor of the Manchu-led Qing dynasty. Photo: SCMP/ Simon Song
Melissa Zhu

Hello, this is Melissa Zhu from SCMP’s tech desk in Hong Kong with a round-up of our top stories in the world of China technology.

This week, we look at how some recent innovations have taken inspiration from history: a video game developer has used artificial intelligence (AI) to recreate the faces of ancient Chinese emperors, while China’s online microlenders are putting a digital twist on the age-old practice of moneylending. And in other news, Chinese regulators have removed TripAdvisor and 104 other apps from app stores for what they have deemed “illegal” content.

Face time

Advertisement

Many great emperors have ruled China over the past 4,000 years, but those of us alive now have only been able to imagine what they might have looked like based on books and ancient illustrations.

Now, AI technology is making it possible to put a realistic face on a number of these historical figures.

Advertisement

Hu Wengu, an independent Chinese video game developer and programmer, has used AI to recreate the animated faces of Chinese emperors from the Song, Ming and Qing dynasties – a period that dates more than 1,000 years – based on historical portraits.

Hu relied on PaddleGAN, an open-source machine learning project by Chinese internet search giant Baidu, and Artbreeder, an online AI tool that mixes and alters images, to recreate the emperors’ faces. In addition, Hu used a system called AI Studio to capture his own facial expressions and apply them to those faces, giving them more lifelike expressions.

Advertisement
Select Voice
Choose your listening speed
Get through articles 2x faster
1.25x
250 WPM
Slow
Average
Fast
1.25x