Why are villagers trading faces for household items?

Question

Accepted Answer

Villagers are trading facial images for household items as compensation for participating in a data collection project to train artificial intelligence software.

In a village in China's Henan province, residents are exchanging images of their faces for everyday household goods such as kettles, pots, and tea cups. This practice is part of a data collection initiative managed by Qianji Data Co Ltd, a company that gathers and annotates data crucial for training AI software. The primary objective of this project is to enable AI systems to differentiate between authentic facial features and still images, with villagers receiving household items as remuneration for their involvement.
Data Collection for AI Training

The project in Henan is specifically designed to collect data that will train AI algorithms, particularly in distinguishing real facial features from static photographs. Liu Yangfeng, CEO of Qianji Data Co Ltd, stated that their company is developing extensive datasets to support numerous AI algorithm companies, thereby contributing to the advancement of artificial intelligence in China. These datasets are essential for AI development, allowing algorithms to learn and improve their performance on various tasks.
Compensation Through Household Goods

Participants in the project are compensated with tangible household items rather than monetary payment. Villagers have been observed lining up, some reportedly for the third or fourth time, to provide their facial images. This method of compensation, involving items like kettles and pots, serves as an incentive for local residents to contribute to the data collection effort.
The Broader Data Annotation Industry

This initiative is indicative of a larger global industry focused on data collection and labelling for AI training. Companies worldwide are involved in gathering and annotating information, such as photos and videos, to teach machines how to interpret visual data. The global market for machine-learning data annotation has seen significant growth, estimated to have increased by 66 per cent to US$500 million in 2018 and projected to more than double by 2023, highlighting the substantial demand for such services.

Reading history

Hide →

This history is cleared when the page is closed.

Why are villagers trading faces for household items?

Loading

Was this article helpful?

Helpful

Not helpful

Report an issue

1

0

0

Share with friends