samee-r10 / image-classification-using-cnn

Convolution Neural Network

Image-Classification-Using-CNN

Convolution Neural Network E-commerce has rapidly grown and their business strategies are completely based on user actions and user experiences. Although it is completely based on users, we should also not forget to say that there is a technology bridge in between users and growth in business. It may be Machine Learning or Deep Learning. Companies apply many image classification techniques on data to improve their catalog and give best suggestions to the users. They need accurate product classification on their platforms for better user experience. But when you talk about products, there exists a huge variety and classifying within varieties is really challenging. As a Deep Learning engineer, you should always try cracking these kinds of challenges by classifying things within a product itself.

Goal: Given the images of a product with multiple categories, train a model which can classify the type of a product.

Data Description: Data is all about images of shoes with multiple categories and data is collected from a popular Ecommerce site. Data set consists of two folders train and test.

Provided Files: Train: train set consists of images belonging to 3 different categories of shoes in 3 differentfolders: Boots, Sandals and Slippers. Test: test set consists of images belonging to all 3 categories of shoes into a single folder.

Instructions:

1. Train set should be used to feed the model.
2. Test set should be used to predict labels for test data. Evaluation Criteria: The evaluation metric for this problem statement is the Accuracy scorewhere each shoe category is matched with the actual shoe label.