Home / Business / Technology

Baidu uses AI technology to help abducted child find family

Xinhua | Updated: 2017-04-13 13:12

Baidu uses AI technology to help abducted child find family

A photo of Fu Gui at 4 posted by his family on (left) and a photo of Fu Gui at 10 posted by himself on the same site.[Photos provided to]   

BEIJING — Baidu succeeded in using artificial intelligence to reconnect a man with his family 27 years after he was abducted, the company announced Wednesday.

Working with, a charity group dedicated to connecting missing children and their families, Baidu used its cross-age facial recognition program to analyze pictures of abducted children uploaded by the victims and their birth families and identified potential matches through the comparison of selected facial features.

Fu Gui, 33, who was born in Chongqing Municipality, western China, was abducted in 1990 and later transferred to Fujian Province, southeastern China. He registered with in 2009, and his natural family did the same in early 2017.

Baidu's facial recognition program was able to draw up a short list of potential identities for the man from pictures uploaded to the site, and a DNA test later verified the correct match.

Baidu has about 200 million sample pictures that it uses to improve the sensitivity and accuracy of its facial recognition program, which can be over 99 percent accurate.

Baidu's founder and CEO Robin Li, who is also a national political advisor, proposed during the annual parliamentary sessions in March that AI systems, like facial recognition, could be used to help find missing children and suggested that there should be a central database with missing children's information.



Most Viewed in 24 Hours
Copyright 1995 - . All rights reserved. The content (including but not limited to text, photo, multimedia information, etc) published in this site belongs to China Daily Information Co (CDIC). Without written authorization from CDIC, such content shall not be republished or used in any form. Note: Browsers with 1024*768 or higher resolution are suggested for this site.
License for publishing multimedia online 0108263

Registration Number: 130349