Finding trafficked children through crowdsourcing

• Main Authors: Goh, Dion Hoe-Lian, Lee, Chei Sian, Guo, Hang
• Other Authors: Wee Kim Wee School of Communication and Information
• Format: Article
• Language: English
• Published: 2021
• Subjects: Social sciences::Communication; Engineering::Computer science and engineering; Child Trafficking; Crowdsourcing
• Online Access: https://hdl.handle.net/10356/151591
• Institution: Nanyang Technological University

Tackling child trafficking requires a multi-pronged approach from government and non-government organizations as well as individuals. In this paper, we introduce Zhongxun, a crowdsourcing platform for finding missing children that have been potentially trafficked. Users submit photos of children whom they suspect may be victims of trafficking. Zhongxun uses a facial recognition algorithm to match the submission against its database of photos missing children and returns those that are similar for followup action. Users may also provide feedback on the matches to improve the facial recognition algortihm. Zhongxun is a live system and is currently being used in China.