CNN
Convolutional Neural Network
An AI technique that excels at picking out features in an image to identify what it shows.
In Simple Terms
A CNN is a mechanism that detects features like shapes and colors in an image to figure out what it shows. Rather than looking at the whole image at once, it repeatedly breaks the image into small regions and extracts features from each one. This approach is widely used in everyday services, such as photo analysis and image search.
Behind the Name
CNN stands for Convolutional Neural Network. The word "convolutional" comes from a mathematical operation called convolution, while "Neural Network" describes a computational model inspired by how neurons in the human brain work.
Take a Closer Look!
A CNN is a computational mechanism that reads out features at specific positions in an image, step by step, to identify what's shown. Through processes like convolutional layers, it gradually extracts features from the image as it computes.
When processing an image, a CNN first picks out simple features like outlines and lines within small regions. Next, it combines the lines it found to recognize more complex shapes, such as eyes and a nose, and finally judges what the whole picture shows. Rather than processing the image exactly as it is, it has the property of recognizing the same feature even if its position shifts slightly.
This computational approach is used in many situations involving images, such as automatic handwriting recognition, medical image diagnosis, and image search services. Since video can also be treated as a collection of images, like a flipbook, it's also applied to technology that analyzes motion and objects within video.