Classification 6 categories of landscape images, taken from Intel's dataset, from a competition hosted on Analytics Vidhya. Written in Python with PyTorch.
Architectures applied: Custom CNNs, Residual Networks
Max accuracy attained: 91%
Transfer learning using Torchvision
In depth explanation at https://medium.com/@devbhartra183/how-to-build-a-landscape-photo-classifier-in-pytorch-cnns-resnet-and-transfer-learning-97682a2a1eb1?source=friends_link&sk=613189b99db8aed10d394b16dfe01dac