Should I used resnet 34?

It is my understanding that resnet 34 is a model that has been pretained on photos. Is it ok to use this model for images that are not really photos?

My dataset is made up of images that consist of a single line making a closed contour (similar to hand written digits in the MNIST dataset). Is there a different model that would be better than resent?


I think it would be a very good start to try a pretrained resnet model and retrain it.

If you have enough data you could also try to train from scratch.