djmv / mobilnet_ssd_opencv Goto Github PK

Hi! We really appreciate your work developing and training this net! Would you be willing to add an open-source license to your code? Preferably Apache, BSD, or similar?

Can please guide me through?

for video input
ModuleNotFoundError: No module named 'numpy'

for image input

File "sample_img.py", line 109
print label
^
SyntaxError: Missing parentheses in call to 'print'. Did you mean print(label)?

I am impressed with your work. I want to start to make another mobilnet SSD. So, how can I improve better than you?

Can you show example how this can be used with full googlenet model?

I use pytorch to get my model about mobilenet-ssd, and I transformed my model（.pth）to onnx model,and onnx to caffemodel,
so I get my caffemodel with not using caffe to train.
But when I use your code to run, I meet a mistake about dim:
In your code, the detections(about net.forward())'s shape is (1, 1, 100, 7), but my shape is (1, 2268 ,3), so I reshape the detections to (1, 1, 972, 3),the code can run ,but It doesn't work well, It can't do anything about confidence and Object location.
So how can I do for the confidence and Object location.
This is my model: https://drive.google.com/drive/folders/1dWObvYe9VRUBJ1D5Teow4Y4M34tJ_6-B?usp=sharing