Academic Projects

List of projects I have completed as a part of various courses taken in college.

  1. Classification and Diagnosis of COVID-19 from Medical Images (May 2021)

    Implemented a supervised Convolutional Neural Network for classification of medical images for detecting the presence of Covid/Pneumonia affected lung images. We noticed that the models used for similar purposes at the time only worked with good accuracy on the training datasets, whereas on a different dataset that contained similar CT Scan images the accuracy was reduced to 60-70%. To improve the robustness of our model we implemented a voting based approach that considers multiple images of the same person as a single group and also performed cross-data analysis. This was done as the project work for the course 19MAT212 :‘Mathematics for Intelligent Systems 4’.

  2. Twitter Data Analysis (May 2021)

    Implemented a LSTM based model for classifying the authenticity of covid related tweets. Dataset containing the tweets was extracted using Twitter API. This was done as the project work for the course 19AIE214:‘Big Data Analytics’.

  3. Compressing Genome sequences using Burrows Wheeler Transformation (December 2020)

    Implemented a python program that can performs BWT on very large genome sequences and run length encoding was done for compression. Test was done on a Coronavirus dataset. This was done as the project work for the course 19BIO201:‘Intelligence of Biological Systems 3’.

  4. Cost Effective Health Monitoring Device (May 2020)

    Proposed a prototype for a wearable health monitoring device particularly useful for people during the pandemic times. This paper was accepted for COCONet conference 2020. This was done as the project work for the course 19AIE114: Principles of Measurements and Sensors.

  5. Applications of Fourier Transformation for Image Processing (April 2020)

    Implemented a MATLAB application that computes FFT and inverse FFT of an input image and performed edge detection and image de-noising. This was done as the project work for the course 19MAT117:‘Mathematics for Intelligent Systems 2’.

  6. MATLAB based GUI application to find the Centroid and Moment of Inertia of any input shape. (December 2019)

    Implemented a MATLAB program that can compute the Centroid and Moment of Inertia of an input shape, a GUI was implemented asking users about the input shape and its dimensions. This was done as the project work for the course 19PHY104:‘Computational Engineering Mechanics 1’.

Personal Projects

Most of the personal project projects I have worked on in the past has been documented in my blog page. Here is a list of some of the recent ones that I have completed.

  1. Siamese Network for Image Classification (May 2021)

    Used a Siamese Network that was originally proposed for one-shot learning application such as face detection/signature verification, for detecting parasitized cell images. [Blog] [Notebook]

  2. Attention based CNN for Image Classification (April 2021)

    Implemented a CNN architecture that made use of attention mechanisms for simple Image Classification application on CIFAR-10 dataset. [Blog] [Notebook]

  3. Open AI Gym Problem (March 2021)

    Implemented a keras sequential model to solve a Reinforcement Learning problem in OpenAI Gym. [Blog] [Notebook]

  4. Implemented StyleGAN model (February 2021)

    Implemented StyleGAN using PyTorch for generating realistic images of human faces. [Blog] [Notebook]

  5. Implemented SinGAN model (January 2021)

    Implemented SinGAN using PyTorch for Image Style Transfer. [Blog] [Notebook]

  6. Implemented CycleGAN model and modified its architecture for faster training(December 2020)

    Implemented CycleGAN model using PyTorch for Image to Image Translation. Improved its architecture and created a new day-night dataset for testing its performance. [Blog] [Notebook]

  7. Implemented SRGAN model and improved its performance (December 2020)

    Implemented SRGAN model using PyTorch for Single Image Super Resolution and made improvements to its loss funtions. [Blog] [Notebook]