samyuen101234 / masked_face_recognition Goto Github PK

I noticed that some of the images in your Casia dataset have purple surgical mask in it. Can you tell me what's the hex color for that?

Sorry, the file you have requested does not exist.

Make sure that you have the correct URL and the file exists.

I'm trying to load the .pt model you provided, but I can't seem to make it work
Code:

model = InceptionResnetV1(pretrained='vggface2')
checkpoint = torch.load("/content/drive/MyDrive/export/InceptionResNetV1_ArcFace.pt")
model.load_state_dict(checkpoint)
model = model.cuda()

Output:

`ModuleNotFoundError                       Traceback (most recent call last)
[<ipython-input-6-9840bef75f84>](https://localhost:8080/#) in <module>()
      1 model = InceptionResnetV1(pretrained='vggface2')
----> 2 checkpoint= torch.load("/content/drive/MyDrive/export/InceptionResNetV1_ArcFace.pt")
      3 model.load_state_dict(checkpoint)
      4 model = model.cuda()

2 frames
[/usr/local/lib/python3.7/dist-packages/torch/serialization.py](https://localhost:8080/#) in find_class(self, mod_name, name)
    873         def find_class(self, mod_name, name):
    874             mod_name = load_module_mapping.get(mod_name, mod_name)
--> 875             return super().find_class(mod_name, name)
    876 
    877     # Load the data (which may in turn use `persistent_load` to load tensors)

ModuleNotFoundError: No module named 'model'`

Can you please tell me what did I do wrong and how to fix it? I'm a beginner so I don't really know what to do here.

Hi there， deos the tesetset include masked faces?

Hi Sam Yuan,

This is Yuvaraj Singh from India, Studying engineering.

I came across your project “Access Control System with Masked Face Recognition”

I could only get your version and got Test and model results.

Was hoping if you could share the whole project with me was very interested in knowing this project.

In my version of project the accuracy is low and am unable to satisfy the spoofing version of it.

It would be of great help if you could help me out with it As soon as possible.

Thankyou,

Yuvaraj Singh

Mail- [email protected]

Hello!
I ran this code, but I had met "collapse" when 1epoch was over. (I edited code...)
When does "collapse" happens? and How can I solve this problem?

Hello，I could not understand the meaning of the eval graph，could you please explain about it such as the meaning of horizontal coordinate and vertical coordinate？

Hi dear,
I have downloaded the Model1 and tended to finetune it. But it raise error that

Traceback (most recent call last):
File "main2.py", line 81, in
facenet, optimizer, scheduler = load(name)
File "/root/maskedface/Masked_Face_Recognition/train.py", line 26, in load
model = checkpoint['model']
TypeError: 'FaceNet2' object is not subscriptable

so how to load your trained model1?
And where is the code of finetune?

We are trying to recreate your execellent work.

We are using facenet = torch.load('../Model/InceptionResNetV1_ArcFace.pt') from the shared gDrive
we are getting error in the line distance = torch.norm(eval_out - eval_pait_out, dim=1)
can you please help in getting over this issue ?

(model-training) C:\Users\ranji\AML-ReferenceCode\Masked_Face_Recognition\Code>python ./main_test.py
Device: cpu
0%| | 0/456 [00:22<?, ?it/s]
Traceback (most recent call last):
File "C:\Users\ranji\AML-ReferenceCode\Masked_Face_Recognition\Code\main_test.py", line 72, in
dist_threshold = evalulate(facenet, eval_loader1, eval_loader2, device)
File "C:\Users\ranji\AML-ReferenceCode\Masked_Face_Recognition\Code\test.py", line 56, in evalulate
dist1 = result(model,eval_loader1, device)
File "C:\Users\ranji\AML-ReferenceCode\Masked_Face_Recognition\Code\test.py", line 42, in result
distance = torch.norm(eval_out - eval_pait_out, dim=1)
TypeError: unsupported operand type(s) for -: 'dict' and 'dict'

Thanks for your good job! I have download the data, but i do not know how to use it. In your code,how to modify it?

Hi,
0000105 folder (in Training Set1) dosen't exist.
Could you upload that data?

In this real faces dataset, there are 75 real people, and each class has three non masked and masked images. Can you test this data set in your code, and how to operate it

Greetings! Thank you for this useful contribution.
How can I try this out? I.e. you provided 3 models, is there a script for the inference? You also provide a link to some gh repo where you used it, but the link is broken. Or maybe you could give me a hint so I can get it running. Since we are talking about face recognition, I assume that somewhere gotta be a part where I build a face embedding and use some distance metric. Is that correct?

I found that the images in the train, eval, test set have the size of 128x128. However, the images in the real masked face dataset have diverse sizes. How to use the real masked face dataset to test the model trained with images of size 128x128?