Facial recognition, a 'convenient' and 'efficient' solution looking for a problem?
summary
Panel gives concrete examples: a commercial app scraping 3 billion images and used by 600 police forces (Clearview); court rulings (South Wales) that limit police use; trials at major stations reporting 0.5% false-hit rate → ~2,300 wrong IDs per 460,000 daily travellers; rejected school access pilot; promising hololens pilot helping a child with dementia; calls for mandatory third‑party testing, dataset “data sheets,” and legal bans or moratoria for public‑space identification under GDPR and law‑enforcement rules.
AI-written summary · about
chapters (5)
- 0:00Start
- 11:37Is Facial Recognition a Convenient or an Efficient Solution Looking for a Problem
- 22:43Biometrics Are Not Perfect
- 23:25False Acceptance Rate and False Rejection Rate
- 56:51Ai Principles for Facial Recognition
related talks
01h 15mHOW TO RECONCILE FACIAL RECOGNITION TECHNOLOGIES WITH CONSUMERS' PRIVACY
54mFast and Loose with Facial Recognition Technology: Modern Tech, Race, and Civil Liberties
40mNew Face, Who Dis? Protecting Privacy in a World of Surveillance
06mFacial Recognition: Understanding Privacy Concerns and Attitudes Across Increasingly
41mYeah, But Should We....
19mGDPR, AI & use of Video data
01h 14mCPDP ACADEMIC SESSION II
51mWe are the Guardians of our Future