What is Computer Vision?
“Artificial Intelligence” is the great umbrella carrying underneath “Machine Learning” and “Deep Learning”. Computer vision is a subset of machine learning aiming to understand features in an image along with deriving useful insights.
Comparison
This is where most people get confused especially when computer vision technologies are introduced such as object detection, image classification and semantic segmentation. Image classification is used when we would like to label the whole image by a single class or multi classes. So, we could have classes such as “dog”, “cat” and “mouse” and the goal is to observe an image and classify it as containing a dog or a cat or a mouse or a mixture of those (multi classes per image). Object detection is slightly more difficult than image classification in which it should it analyze an image, predict all the different classes present in an image with a confidence score for each class and draw a boundary box around the classes. On the other hand, semantic segmentation provides a much deeper view of the image than object detection. It classifies each pixel in the image to the corresponding class rather than just framing the object, so you get to know more about the dimensions of the object.
For clearly distinguishing between the three algorithms, follow the flowchart below:
Top comments (0)