You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Implemented fully-connected DNN of arbitrary depth with Batch Norm and Dropout, three-layer ConvNet with Spatial Batch Norm in NumPy. The update rules used for training are SGD, SGD+Momentum, RMSProp and Adam. Implemented three block ResNet in PyTorch, with 10 epochs of training achieves 73.60% accuracy on test set.
DogBreedSpotter is a Python-based image classification project designed to identify and classify dog breeds in images. This project utilizes deep learning models, including convolutional neural networks (CNNs) such as VGG, AlexNet, and ResNet, to accurately detect whether an image contains a dog and, if so, determine the breed.
Photo sharing and photo storage services like to have location data for each photo that is uploaded. With the location data, these services can build advanced features, such as automatic suggestion of relevant tags or automatic photo organization, which help provide a compelling user experience.