学进去

Human memory is notoriously (众所周知地) unreliable. Even people with the sharpest facial recognition skills can only remember so much.

It’s tough to quantify how good a person is 1． remembering. No one really knows how many different faces someone can recall, for example, but various estimates tend to hover in the thousands – based on the number of acquaintances a person 2． have.

Machines aren’t limited this way. Give the right computer a massive database of faces, and it can process what it sees – then recognize a face it 3． (tell) to find – with remarkable speed and precision. This skill is 4． supports the enormous promise of facial-recognition software in the 21^st century. It is also what makes contemporary surveillance (监控) systems so scary.

The thing is, machines still have limitations when it comes to facial recognition. And scientists are only just beginning to understand what those constraints are. 5． (figure) out how computers are struggling, researchers at the University of Washington created a massive database of faces – they call it MegaFace – and 6． (test) a variety of facial-recognition algorithms(算法) as they scales up in complexity. The idea was to test the machines on a database that included up to 1 million different images of nearly 7,000 different people – and not just a large database 7． (feature) a relatively small number of different faces, more consistent with what’s been used in other research.

As the databases grew, machine accuracy dipped across the board. Algorithms 8． were right 95% of the time when they were dealing with a 13,000-image database, for example, were accurate about 70% of the time when 9． (face) with 1 million images. That’s still pretty good, says one of the researchers, Ira Kemelmacher-Shlizerman. “Much better than we expected,” she said,

Machines also had difficulty adjusting for people who look a lot alike –either doppelgangers (长相极相似的人), whom the machine would have trouble   10． (identify) as two separate people, or the same person who appeared in different photos at different ages or in different lighting, whom the machine would incorrectly view as separate people.