shape problem

[seogihyun]

I got a model_baddy.h5 file from weekenddeeplearning because your model.1 file does not operate.
However there is other problem "Cannot assign to variable conv2d/kernel:0 due to variable shape (3, 3, 3, 64) and value shape (64, 9, 3, 3) are incompatible".

I want to fix the shape.
Please let me know.

[Yasminbr]

Hey im isuuimg the same problem in my code too .

[Q2Learn]

I had the same problem and I had to check all the file names because some were Capital and some we Non Capital. Also some folder names had a '/' in front of it so if you're in working directory and not root directory than make sure your directory location is good as well as the file and folder names.

For example some files were Yolov3 instead of the yolov3 and '/VideoOutput' instead of 'video_output':

After changing my files and folders I have it working with:

!python3 "predict_video.py" --save_weights_path="weights_tracknet/model.1" --input_video_path="VideoInput/video_cut.mp4" --output_video_path="VideoOutput/video_output_MyTest.avi" --n_classes=256 --path_yolo_classes="yolov3/yolov3.txt" --path_yolo_weights="yolov3/yolov3.weights" --path_yolo_config="yolov3/yolov3.cfg"

Also check in file predict_video.py on line 66 that it's 'modelFN = trackNet' and not 'modelFN = tracknet'

[Gi-gigi]

Hi @Q2Learn friend, I updated my path, but also ocurred:

`Traceback (most recent call last):
  File "tennis_tracking/predict_video.py", line 109, in <module>outs = net.forward(get_output_layers(net))
  File "tennis_tracking/TrackPlayers/trackplayers.py", line 10, in get_output_layers output_layers = [layer_names[i[0] - 1] 
for i in net.getUnconnectedOutLayers()]
  File "tennis_tracking/TrackPlayers/trackplayers.py", line 10, in <listcomp>
    output_layers = [layer_names[i[0] - 1] for i in net.getUnconnectedOutLayers()]

IndexError: invalid index to scalar variable.`

can u help me how to fix it? thank u~

[Q2Learn]

Hey,

I’m not sure. I haven’t worked with this algorithm in a while. And I’m not familiar with it. But check all your paths are correct in the code base because it’s looking for the model weights in specific locations. That’s how I got it to work anyhow.

[Gi-gigi]

Hey friend @Q2Learn.
One more question, do u run this project on google colab? or personal PC?