Facial Recognition

Facial recognition tests, iris, speech pattern recognition & fingerprint analysis are real life uses of pattern recognition in machine learning. Read More

Computers are extremely good at finding complex patterns in data. The application of machine learning allows pattern recognition programs to create algorithms (a set of instructions) to notice complex patterns in numbers, text, and pictures. The program does this by analyzing specially created data known as training data to abstract general rules and procedures. These new rules and procedures are tested on new data to determine if the “abstraction” worked properly or if the programmer needs to change the rules and apply the training data again. Learning also means that over time as the programs experience grows it becomes better at recognizing patterns.

Recognizing faces is grouped in a field of pattern recognition known as biometrics.  Biometrics is a form of identity through the recognition of unique individual characteristics. These include fingerprints, DNA, face structure, the iris of our eyes or the retina at the back of our eyes, and speech recognition. For use in identification, a biometric must be shared by all or the majority of people (e.g. all people have fingerprints), be unique for each individual, have minimal change over the lifetime of an individual, be easy to collect, and not be easily duplicated or faked.

Did you know that biometric systems are used to secure buildings and important, secret information either instead of a password or used together with a password for added protection? Biometrics can also be used to find a particular person in a large crowd of people. You can learn more about machine learning by using the experiments and trying the interactive fact sheet available at wonderville.ca.

Facial recognition tests, iris, speech pattern recognition & fingerprint analysis are real life uses of pattern recognition in machine learning.