7
❝
❞
AI and ML Technology
Take an About-Face
WHAT'S NEXT WITH VOICE CONTROL
By Paul Golata, Mouser Electronics
F
rom our earliest days in infancy, human beings possess
the amazing ability to seek out and distinguish faces.
This capability needs no external training; it develops
naturally based on instinct and repeated exposures.
Even under varying circumstances such as lighting, angle,
expression, or changes such as facial
accessories, humans are able to quickly
recognize faces. Let's examine how Artificial
Intelligence (AI) and Machine Learning (ML)
technologies support facial recognition
applications.
Unlike the human brain, contextual variabilities
make facial recognition a significant
challenge for computer vision. Fortunately,
the advancement in high resolution and 3D
imaging capabilities over the past decade,
combined with the use of convolutional
neural networks (CNNs) for model training,
significantly improved the accuracy of the facial
recognition technology today, allowing facial
recognition to be implemented in applications
ranging from home appliances to smart access
control for secure industrial environments.
Inference latency time requirements
determine the best MCU or MPU for
your facial recognition application.
AI
