Image Classifier to recognize the environment of the image

Over this project we are going to utilise the Intel image Classification dataset from kaggle. More about the dataset is explained below.

We would be using python's torch module for pre-processing, training and also for inference. This project would utilize convolutional neural network as it would give more spatial understanding to the model. Deep architectures usually find it difficult to produce meaningful data deeper in the netwrok due to vanishing gradients thus this model makes use of residuals.

This model resembles the structure of Resnet9 architecture.

import os
import torch
import torchvision
import tarfile
import torch.nn as nn
import numpy as np
import torch.nn.functional as F
from torchvision.datasets import ImageFolder
from torch.utils.data import DataLoader, TensorDataset
import torchvision.transforms as tt
from torch.utils.data import random_split
from torchvision.utils import make_grid
import matplotlib
import matplotlib.pyplot as plt
%matplotlib inline

project_name = 'course-project'

Image Classifier to recognize the environment of the image

The data set is from kaggle and can be viewed here.