-> Architecture of ResNet18 was cusomized for our use case.
-> Feature extraction from last fc layer of Resnet18 for Train and Test images in Datasets.
-> Classification using k-Nearest Neighbours.
-> Accuracy Check (k=5) : 95%
-> Architecture of ResNet18 was cusomized for our use case.
-> Last fc layer was replaced with new fc layer for the number of classes in our dataset.
-> Trained the model on training dataset.
-> Accuracy Check for epochs = 100 : range between [ 97.15, 100 ]