- Task is to classify the images as live or fake
- Data inputing:
- Using torchvision.datasets module from torchvision library we converted raw images of fingerprints into tensors and given as input to the model
- To achieve faster training of model, by using torch.utils, converted the data into dataloader object.
- Loss function used is crossEntropyLoss and the optimizer used is stochastic gradient descent.
- Number of classes involved in classification are 2(live, fake)
- Number of epochs are taken to be 10.
- We got an accuracy of 86%
- Accuracy can be improved by increasing the size of data set and number of epochs
Some of the output images predcted by the model, view from the implementation file