Wolfram Computation Meets Knowledge

Wolfram Summer School

Alumni

Harish Chetty

Technology and Innovation

Class of 2017

Bio

I am currently pursuing my bachelor’s in information technology (IT) from NMIMS University, India. I have a great interest in technology and software development. I have worked on image processing as a research topic and am currently working on data encryption and decryption. The concept of the Internet of things (IOT) is really interesting and I am learning more about it. I was excited to learn and work with the Wolfram Language and enjoyed doing that this summer. In my free time I like to write poems, discuss philosophy and listen to music.

Computational Essay

Laplace Transform »

Project: Leaf Classifier

Goal of the project:

The goal of the project is to classify leaves on the basis of their species using a convolution neural network (CNN) model.

Summary of work:

A dataset of different species of leaves from the northeastern United States and Canada was taken and was split into training and validation sets. A convolutional neural network (CNN) model was developed for the application of deep learning in image processing. The network was trained on this dataset and a leaf classifier was developed.

Results and future work:

The network is able to classify the leaf on the basis of its species most of the time. With more training and parameters, the accuracy level of the classifier can be improved. This network can be useful on a larger dataset with species of leaves from around the world, and thus has a vast range of possibilities to classify from. With more training and a large variety of datasets, a lot can be achieved in the future.