I’m trying to check if the trained model is more generalized version than the previous models.
Are there any techniques, or Tools for seeing if the model is enhanced in terms of generalization? (Model generalizability for image classification)
The usual way to do this is to keep a subset of your data for validation: you train both model on a training subset of your data. After training, you evaluate both on this held out validation set to see how they generalize to this unseen data.