What is the difference between image classification and image recognition (in the context of face recognition)?

Question

A classic example of image classification problem is to classify handwritten digits using softmax linear regression model for MNIST data. Let us suppose there is a facial database of 10 subjects and 10 images for each subject. This will be a problem of image (face) recognition.

So considering the difference in the feature space of digits and faces, by analogy can I assume each subject as each digit and images of subjects as examples of handwritten digits and apply the classification algorithm to perform recognition. Check this link

Please help me understand this?

Answer 1

In this context there is no difference. Just two data sets for a classification problem: handwritten digit images with digits as labels or face images with subjects as labels.

However normally face recognition tasks involve finding all faces in one picture, not labeling them

Answer 2

In the context of face recognition when you say recognizing you are identifying ROI as person out of image while in classification you classify ROI to one of predefined class say male or female.