purehing / face-mask-detection-tf2 Goto Github PK

Hi, I would use with GPU but I have problem with Jetson Nano.
Is it necessary setting something to work with GPU and tensorflow?
I am using the trained model.

Thansk
Regards

Thanks for the wonderful project!

In addition, it would be useful if you could specify a type of LICENSE file, depending on how you want to give access to others to use this project.

The code is great and straightforward.
I just thought to make it more popular just:

1 - add the pretrained models (just like you shared the data).
2 - Give recommendation on how we probably can boost up the performance. (like changing the kernels, change the backbone or etc ...)

Thanks

I installed the requirements.txt, then ran

python inference.py  --model_path checkpoints/ --img_path assets/1_Handshaking_Handshaking_1_71.jpg

encountering errors as following:

[ WARN:0] global /io/opencv/modules/videoio/src/cap_v4l.cpp (887) open VIDEOIO(V4L2:/dev/video0): can't open camera by index
No camera found
'NoneType' object has no attribute 'shape'

What's the reason? Thanks in advance.

1. masked
2. unmasked
3. background
   is it right?
   thank you!

Hi

I have tried both nncase0.2 and 0.1 with no luck with MaixPy.

Here are the outputs
when trying to convert with nncase 0.1 it gives relu not supported. I changed that to leakyrelu because their documentation says leakyrelu is supported. But then I get an exception in nncase:

   Fatal: Nullable object must have a value.
   System.InvalidOperationException: Nullable object must have a value.
   at System.Nullable`1.get_Value()
   at NnCase.Converter.Converters.TfLiteToGraphConverter.ConvertReshape(Operator op) in D:\Work\Repository\nncase\src\NnCase.Converter\Converters\TfLiteToGraphConverter.cs:line 197
   at NnCase.Converter.Converters.TfLiteToGraphConverter.ConvertOperator(Operator op) in D:\Work\Repository\nncase\src\NnCase.Converter\Converters\TfLiteToGraphConverter.cs:line 97
   at System.Linq.Enumerable.SelectEnumerableIterator`2.ToList()
   at System.Linq.Enumerable.ToList[TSource](IEnumerable`1 source)
   at NnCase.Converter.Converters.TfLiteToGraphConverter.Convert() in D:\Work\Repository\nncase\src\NnCase.Converter\Converters\TfLiteToGraphConverter.cs:line 34
   at NnCase.Cli.Program.Main(String[] args) in D:\Work\Repository\nncase\src\NnCase.Cli\Program.cs:line 113
   at NnCase.Cli.Program.<Main>(String[] args) 

With nncase 0.2 the model converts but it doesn't load under maixpy, give following error:

ValueError: [MAIXPY]kpu: load error -3

basically kpu.load() lead to this error. From the documentation I couldn't find what -3 means.

I tried reducing layers in case it is a size issue, but still the issue persists.

Next I will try without maixpy, using their SDK, that will take some time.

I am sure the problem is with nncase and k210 not with this code or model. Opened a ticket on their maixpy repo as well, lets see what they say about it.

How to choose the min size, steps, match_thresh and variance in the anchor settings?

This is my output from running kmeans.py on your maskdata dataset

8,6 , 18,15 , 30,24 , 35,33 , 52,44 , 54,50 , 76,67 , 79,67 , 103,88 , 108,92 , 111,114 , 137,127 , 139,134 , 156,156 , 187,180 , 193,232

I tensorflow/core/platform/cpu_feature_guard.cc:141] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2
Dimensions must be equal, but are 52960 and 4420 for 'mul_1' (op: 'Mul') with input shapes: [52960,2], [4420,2].

i am getting this error while running inference on camera.

What are x_scale, y_scale, width_scale, height_scale values? Also, what is the number of bounding boxes?

Hi,
I would test this code but I have problem to connect to Baidu link to download your revision.
Can you give me other link to download?
My mail is [email protected]
Thnaks for your availability

Thank you for good repo.
I want to convert h5 to kmodel. but it doesn't work on tf2.1
How did you convert h5 file to kmodel?

Hi, I tried to run tensorflow_infer.py downloaded from AIZOOtech but I dad problems with tensorflow_loader.py.
Did you change something in code?
I have tensorflow 2.1 version with Python 3.7.7.
I have not GPU
Can you help me?
Thanks

Hi, thanks for such a wonderful link. but i am not able to download the dataset from given link. please help or upload it google drive.

thank you.

I tried to run on video with this command

python inference.py --model_path checkpoints/ --camera True

but it give me the following error

W tensorflow/stream_executor/gpu/redzone_allocator.cc:314] Internal: Invoking GPU asm compilation is supported on Cuda non-Windows platforms only
Relying on driver to perform ptx compilation.
Modify $PATH to customize ptxas location.
This message will be only logged once.

any solution?

Hi

I have a question...
How can i convert data to kmodel for K210?

I want to run it on the M5StickV ...

thanks
gadjodilo

Hello thanks for ure project , could you please tell me how can be detect face mask for street view many thanks

After training, I've converted the model to a tflite format and tried to convert it to a kmodel using the following command

~/nncase/ncc compile /home/sean/face-mask-detection-tf2/model.tflite yolo.kmodel -i tflite -o kmodel -t k210 --dataset /home/sean/face-mask-detection-tf2/Maskdata/JPEGImages

But when running the model, I get the following errors

1. Import graph...
   Fatal: Not supported tflite opcode: SHAPE

Hi,
I would change resolution of captured video.
I see that now it is 320x240 in the interference.py.
I tried to change to 640x480 but I had problems in the run.
Can you help me to do this?
Thanks