Be aware of the current trends, terminology, and resources associated with facial recognition.

Research is being performed by universities and private industry in the Biometrics field. You should be aware of the following facial recognition resources:

General Resources

• You meet facial recognition vendors at the annual Biometric conference. To learn more about this conference click here.
• There are many conferences. To view this list click here.
• A resource for learning more about facial recognition can be found by clicking here.
• The methods of matching faces starts with an EigenFace. A quick reference and definition can be found at by clicking here.
• One subtlety of facial recognition is similarities of faces in large populations. An excellent resource to describes this challenge can be found at this link.

Limitations

Be wary of the pitfalls of facial recognition. The engines have shown weaknesses and bias. Issues include: lighting, pose, aging, smiling, ethnicity and familiarity. Clearly, the technology still has a way to go.

• The problems of ethnicity and lighting in one application can be found by clicking here.
• Smiling can be an issue. Should it be? Check out this article about smiling and facial recognition by clicking here.

Studies by psychologists show that humans can match and even outperform computers in facial recognition for people they know, but are not as good for people they do not know. This makes sense as humans have the chance to study and learn faces from several angles. We have developed the means for computers to learn as easily.