Colorado university professor secretly films campus students to improve facial recognition models

A University of Colorado professor secretly photographed students, faculty members and others walking in public in order to improve facial recognition systems, prompting accusations of privacy abuse and poor ethics. Professor Terrance Boult secretly photographed thousands of campus students, employees and visitors on public sidewalks at the university's Colorado Springs campus in 2012 and 2013, using a hidden camera set up in a building nearly 500 feet away. The aim was to improve facial recognition, especially for long-range or surveillance applications. In 2016, Boult posted the Unconstrained College Students (UCSD) online as a publicly downloadable dataset of people acting naturally in public . Of the 2,400 people captured, more than 1,700 were considered “matched identities” - meaning the same person was snapped more than once - and of good enough quality to be used in the study. The project, which r eceived funding from a number of US intelligence and military operations, including the Office of Naval Research, Special Operations Command, and the Office of the Director of National Intelligence , also included a competition for researchers to see if they could correctly match the faces using the developing software. The project was uncovered by the Colorado Springs Independent in May 2019 , a month after the dataset had been taken down. The expose ignited a firestorm amongst the local media, which focused on its opacity and intrusiveness. ➕ September 2019 . T he Financial Times cited the project as an example of an attempt to gather personal images to improve facial recognition systems in as natural (ie. 'wild') a manner as possible - ideally covertly. System 🤖 Unconstrained College Students (UCSD) dataset Operator: Beckman Institute; Beihang University; Inception Institute of Artificial Intelligence, Abu Dhabi; Pontificia Universidad Católica de Chile; Queen Mary University of London; University of Notre Dame; Vision Semantics Developer: University of Colorado Country: USA Sector: Education Purpose: Train facial detection and facial recognition systems Technology: Database/dataset; Facial recognition; Computer vision Issue: Ethics/values; Privacy; Tr ansparency

Date it happened

2019-05-01

Organisation involved

Beckman Institute; Beihang University; Inception Institute of Artificial Intelligence, Abu Dhabi; Pontificia Universidad Católica de Chile; Queen Mary University of London; University of Notre Dame; Vision Semantics

Product, system or model

Unconstrained College Students (UCCS)

This is a record of what was reported, not a finding that anyone broke the law. If it names your organisation and you believe it is wrong, the corrections process is free and open to everyone.