Introduction - If you have any usage issues, please Google them yourself
Face recognition is a supervised learning process. Firstly, a face model is constructed by training set, and then the test set is matched with the training set to find the corresponding training set head. The easiest way is to directly use the Euclidean distance to compute the distance between each image of the test set and each image of the training set, and then select the nearest image as the result of recognition. This method of calculating distances directly is intuitive, but there is a very big flaw - too much computation. If the size of each image is 100*100, and the training set size is 1000, then it is necessary to recognize an image in the test set, and the computation amount of 1000*100*100 is required. When the test set is large, the recognition speed is very slow.