Faces Change With Time: New Test Shows Machines Struggle
Photos of same person at different ages can trick computer systems, so researchers made a new photo set to check that.
A well known database seemed almost perfect, but it didn't pay enough attention to cross-age photos.
To fix that they built CALFW, adding many pairs of the same person taken years apart, to force the system to face real world aging changes.
They also picked mismatched pairs that share gender and race so the test focuses on the faces not on other clues.
The result was clear, machine scores fell a lot — about a 10–17% drop compared to old tests.
This new test shows that time on your face matters, and what worked before may not work in the street or phone photos.
It means builders must train systems to handle years, not just perfect poses.
People will age, photos will change, and tech must learn to keep up, otherwise mistakes will keeps happening.
Read article comprehensive review in Paperium.net:
Cross-Age LFW: A Database for Studying Cross-Age Face Recognition inUnconstrained Environments
🤖 This analysis and review was primarily generated and structured by an AI . The content is provided for informational and quick-review purposes.
Top comments (0)