Do you want your online photos to be used to train computers?

March 27, 2019 10:35
Flickr became an appealing resource for facial recognition researchers because many users published their images under 'Creative Commons' licenses. Photo: Freepik

Face recognition technology has advanced rapidly in recent years. It has become a key part of identifying a person using biometric technology, which can be used to unlock mobile phones, track employee time, identify regular customers in retail shops or even help police to track down criminals.

Still, the technology has some way to go and is yet to be perfected.

It’s said that computers are less good at recognizing people with darker skins, and that machines often mix up genders. Given this, the training database should perhaps add more images of personalities such as black women or men with long hair.

Personally, I have mixed feelings toward these new technologies, especially in relation to how data is collected to train computers.

Online face images are usually downloaded by program developers to train the algorithm, without taking permission of the subjects.

IBM in January released a collection of nearly a million photos that were taken from the photo hosting site Flickr to train the AI, an NBC report said.

IBM is not alone in using publicly available photos on the internet in this way. Many research organizations have collected online photos for training facial recognition systems, with some on much bigger scale.

It might be fairly convenient to use public photos from the Web, but there might be some moral issues.

I believe most people might feel surprised or disconcerted if they learn that their photos are being used without their consent. But if they do not know that their photos are being used for such purposes, they simply have no chance to oppose or ask for their photos to be removed.

AI Now Institute studies the social implications of artificial intelligence. They argue that even though people are willing to share their photos on the internet, it does not necessarily mean approval of unlimited access without being notified, or for whatever purposes.

Flickr has become an appealing resource for facial recognition researchers because many users published their images under “Creative Commons” licenses, which means that others can reuse their pictures without paying license fees.

An Australian photographer said he uploaded his photos to Flickr under Creative Commons licenses to allow nonprofits and artists to use his photos for free. But he certainly does not want his more than 700 images to be swept up for facial recognition technology development.

This article appeared in the Hong Kong Economic Journal on March 26

Translation by Julie Zhu

[Chinese version 中文版]

– Contact us at [email protected]


Venture Partner of Sequoia Capital China, former head of the data committee and vice president at Alibaba Group.