Face Recognition Technology
Author: huifan Time: 2020-09-21
Face recognition is a biometric algorithm, which is a method of identifying or verifying a person's
identity through a face. In addition to recognizing faces, face recognition can also recognize photos,
videos, and dynamic people.
The face recognition data is 100% correct, and there are certain errors. Even if people often say that
seeing is believing, it is also deceiving and has errors. When facial recognition software recognizes
African Americans, blacks and other minorities, it will misidentify or fail to recognize them, but facial
recognition algorithms are constantly changing Updates and technology is not perfect, I believe that in
the future, facial recognition software and recognition algorithms will be able to adapt to all groups.
In addition, facial recognition api has begun to be used to protect some groups of voice workers in a
targeted manner. With the wide range of applications and the characteristics of universality, facial
recognition will become more and more human Transformation is getting closer to people’s lives.
In addition to protecting voice groups and recognizing dynamic faces, facial recognition can also
track a person's activities around the world. This requires the person's face to be recorded by
the system. As long as there is a camera, the facial recognition system will monitor To this
person's movement trajectory. This real-time facial recognition algorithm, in addition to the
daily dynamic attendance check-in, has been used in sports events in the United States.
How does facial recognition works
The face recognition system is based on a computer algorithm to select the specific and
unique face information details in the database. After the selection is completed, these face
details are selected (for example, the distance between the eyebrows and the eyes, the face
carousel, the chin to the lips , The distance between the tips of the nose, etc.) are coded
as digital representations used by computers. And call these digital representation data as
Some facial recognition system
Facial recognition systems have different recognition capabilities under different
conditions. These conditions include poor lighting, poor image resolution, and
poor viewing angles.
When there is a recognition error, we have to think about two key concepts:
False negative means that the facial recognition system cannot correctly match
the facial information of the person with the image in the facial system database.
The result of the inability to correctly match is that the system feedbacks us that
the system has no such person, and the query result is zero.
False alarm means that the facial recognition system matches the facial information
of person A with the facial information of person B in the system database, but the
facial information of the two persons does not match, the match is incorrect,
and the match fails.
When we study facial recognition systems, it is important to observe the
"false positive" rate and "false negative" rate.
Youtube: Huifan Technology
Linkedin: Huifan Technology