https://github.com/hatoonguls/deep-learning-with-pytorch

Deep learning with The Street View House Numbers (SVHN) Dataset
https://github.com/hatoonguls/deep-learning-with-pytorch

Last synced: 7 days ago
JSON representation

Deep learning with The Street View House Numbers (SVHN) Dataset

• Host: GitHub
• URL: https://github.com/hatoonguls/deep-learning-with-pytorch
• Owner: hatoonguls
• Created: 2024-07-20T19:18:03.000Z (6 months ago)
• Default Branch: main
• Last Pushed: 2024-07-20T19:36:39.000Z (6 months ago)
• Last Synced: 2024-11-12T07:35:50.506Z (2 months ago)
• Language: Python
• Homepage:
• Size: 57.6 KB
• Stars: 0
• Watchers: 2
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        

In this project I tackled the problem of classifying images.

Dataset Citation:

Yuval Netzer, Tao Wang, Adam Coates, Alessandro Bissacco, Bo Wu, Andrew Y. Ng Reading Digits in Natural Images with Unsupervised Feature Learning NIPS Workshop on Deep Learning and Unsupervised Feature Learning 2011.

The dataset is called Street View House Numbers (or SVHN). The dataset consists of images of house numbers, taken from Google Street View. There are 10 classes, one for each digit (0-9).

There are 73,257 sample digits for training, 26,032 samples for testing, and 531131 extra samples.

I used the dataset to train 3 different models:

Logistic Regression

Feedforward (Fully-Connected) Neural Network (FNN)

Convolutional Neural Network (CNN)

And analyze the difference between these models by looking at the test accuracy and loss.