This project is tending to make deep learning easy learning and fast deployment!
・Guide book: EN | JP
・Build enviroment: EN | JP
・UI configuration: EN | JP
・Configuration detials : EN | JP
3.1 Install Pytorch from here
3.2 Install the other packages by runing this command:
pip install -r requirements.txt
3.3 Run application by runing this command:
python main_gui.py
3.4 Adjust gui by your customed dataset and have fun!
Now you can see the heatmap of the model prediction just on the GUI!(This heatmap is made by the full connection layer of the model)
You just need to select Heatmap Viewer in Mode, and select fill the configuration items in the Heatmap Viewer, the press the Run button. There will be a Heatmap Viewer window pops up and you can check every heatmap of each image in the folder you just selected.
Because ONNX models will not saved when you train or prune the model, so a new function is added to export the inference model to ONNX format.
This project is based on Pytorch
Thanks Naufal for the gui application development!