nikhilmaguwala / image_classification_resnet50

Image Classification Using Resnet 50

We have used Resnet 50 for image classification.

Built Basic Classifier with MATLAB.

The network has an image input size of 224-by-224-by-3.

Usage

This Requires MATLAB (R2018b and above) and the Deep Learning Toolbox.

This NN have three functions:

• resnetLayers: Creates an untrained network with the network architecture of ResNet-50
• assembleResNet: Creates a ResNet-50 network with weights trained on ImageNet data
• Predict: Demonstrates how to classify an image using a trained ResNet-50 network

Example Predication Output:

Deep Network Analyzer:

Accuracy and Loss Graph