Soham_Roy

Soham Roy

I am an undergraduate student pursuing a bachelor's degree in Electronics and Telecommunication Engineering. I am an electronics enthusiast myself, with interest originating from current electricity and circuits learnt in school. My interests broadly lie in image processing and recognition, machine learning, signal processing and circuit designing. I have completed two projects on Image processing and recognition and one on communication. I have an ongoing project on signal processing (improving the popular Recursive least square adaptive algorithm).

My first project was based on circuit component localization from hand-drawn offline circuit diagrams. I created a model which uses morphological operations to localize circuit components on offline hand-drawn circuit diagrams. A paper on the project was subsequently published in an IEEE conference, Calcon 2020.

My second project was based circuit component recognition from hand-drawn offline circuit diagram. It mainly dealt with the classification of most commonly used circuit components/symbols from its hand drawn image by exploiting the texture-based and shape-based features of the image.In the proposed model, the
pre-processed images of circuit components are used for training and testing a recognition model using a feature set consisting of a texture based feature descriptor, called histogram of
oriented gradients (HOG), and shape based features that include centroid distance, tangent angle, and chain code histogram. In addition, the texture based feature, being large in number
compared to others is optimized using a feature selection algorithm called ReliefF. Classification of components is done by using sequential minimal optimization (SMO)
classifier.

I also did a project on communication using RF Energy Harvesting. The Proposed model employs an accumulate and forward scheme in a wireless Half Duplex Communication Network with Multiple sources with multiple antennas, along with multiple relays and uses optimized relay and source selection schemes to increase throughput.

I am also currently involved in an ongoing project based on improving the traditionalal recursive least squares algorithm using a propotionate gain term and zero attractor.

More details about me can be found on my LinkedIn profile. https://www.linkedin.com/in/soham-roy-1b3055197

Projects

Quick Grader

.Django, NLP, django-rest-framework, React js

Skills

Machine Learning
Embedded Systems
Circuit Design
Signal Processing
Digital Image Processing