Classify SVHN images using pretrained VGG16 with PyTorch
VGG16 is a state-of-the-art model that won ILSVRC (Imagenet Large Scale Visual Recognition Challenge) in 2014. VGG (Visual Geometry Group) is a deep convolutional network that can be used for computer vision tasks like image classification. VGG also offers multiple variants. In this video, I used pretrained VGG16 for classifying SVHN (Street View House Numbers) dataset using PyTorch deep learning framework.
GitHub address: https://github.com/randomaccess2023/M...
For more theoretical details on VGG16, check this out: https://www.geeksforgeeks.org/vgg-16-...
00:54 - Import the required libraries
04:01 - Define the device
05:53 - Apply transformation
08:17 - Load SVHN dataset
17:11 - Define dataloaders
19:27 - The pretrained network (VGG16) architecture
20:17 - Visualize the pretrained model (VGG16) architecture
22:54 - Modified pretrained network (MODIFIED_VGG_16) architecture for SVHN images
27:23 - Visualize the modified pretrained model (MODIFIED_VGG_16) architecture
28:59 - Define the train batch
30:49 - Define the test batch
31:50 - Define the accuracy function
34:22 - Define the model_to_train (MODIFIED_VGG_16), criterion (loss function) and optimizer
35:27 - Train the model
45:32 - Plot model loss and accuracy
50:07 - Model prediction
#datascience #python #pytorch #jupyternotebook #vgg16 #pretrainedmodels #imageclassification #svhndataset #deeplearning #transferlearning