Validation Accuracy of Siamese CNN not increasing, stucks at 0.50

Asked Jul 26 '21 at 22:34

Active Jul 27 '21 at 23:27

Viewed 81 times

I'm training a Siamese CNN model based on X-ray dataset and it gets stuck at 0.50 validation accuracy. Moreover, training and validation loss decrease while training accuracy hovers around 0.50 as well.

I think that is a error or something, but I don't know exactly what it is. Maybe use Cross-Validation could help.

Code:

Archictecture of CNN

Summary

Training the model...(results)

edited Jul 27 '21 at 23:27

asked Jul 26 '21 at 22:34

leo-freecss

What is the question? If it is how to improve your accuracy, that's pretty abstract. The question should be more specific. – Pelicer Jul 27 '21 at 02:47
As far as I know, when it happens it means the neural network isn't learning, so I'd like to know if this is a mistake or if I can improve accuracy. – leo-freecss Jul 27 '21 at 23:17
What is the train/val size? – nithish08 Jul 28 '21 at 00:46

Validation Accuracy of Siamese CNN not increasing, stucks at 0.50

0 Answers0