April 2020
Intermediate to advanced
438 pages
12h 2m
English
If you run the preceding code snippets and plot how the accuracy and loss changes for the training and validation image dataset over the number of epochs, you will get an output as shown in the following screenshot:

This time, 45 test images (out of 300 images) were classified wrongly (that is, the predicted label ≠ ground-truth label) by the fine-tuning model; a few of them are shown in the following screenshot. The prediction accuracy was increased:

Only the weights of the convolutional layers were loaded for the pretrained ...
Read now
Unlock full access