face mask deep learning label display

here is my code. I deep-learned mask recognition using an existing face recognition model, automating only half of the face to be labeled here. However, when displaying on the webcam, I want the masked person 'mask' and the non-masked person to be tracked so that 'no mask person' is boxed as the face size as the face size. How to write code for this?

Answer 1

[yanqi liu]

yes，sir，may be the consuming time too much，may be use face-api.js to get the faster application，such as

https://github.com/justadudewhohacks/face-api.js/

[주영 이]

https://github.com/kairess/mask-detection

I want to use this model. Is this model okay? If it's ok, can I download this model via caffe model importer? I'm not sure what code to write to download.

[yanqi liu]

yes，sir，may be can not load in MATLAB，may be use the http server to use the model in python，and make MATLAB client to develop

[yanqi liu]

yes，sir，please use

https://github.com/matlab-deep-learning/COVID19-Face-Mask-Detection-using-deep-learning

the result as follow

I = imread('./SampleMaskData/demo.jpg');
resz = [224 224];
Ir = imresize(I, resz, 'bilinear');
[bbox,score,label] = detect(detector, Ir, 'threshold', 0.5);
bboxf([1 3]) = size(I,2)/resz(2)*bbox([1 3]);
bboxf([2 4]) = size(I,1)/resz(1)*bbox([2 4]);
detectedImg = insertObjectAnnotation(I, 'rectangle', bboxf, [string(label)+ " : "+string(score)], 'Color', 'magenta', ...
    'Fontsize', 30, 'linewidth', 30, 'textboxopacity', 0.4);
detectedImg = insertText(detectedImg, [10, 1],  "    Find mask       ", 'FontSize', 20, 'BoxColor', 'g');
figure
imshow(detectedImg)