About me

I am a 3^rd-year Computer Science undergraduate at IEM, Kolkata, and I am strongly interested in Theoretical Computer Science and Cryptography. I am also interested in applications of Machine Learning in everyday problems.
Currently, I am a Long-Term Project Student under Prof. Avishek Adhikary at Presidency University, working with integrating segmentation and inpainting techniques from computer vision with Secret Image Sharing.
Looking ahead, I aspire to pursue higher studies, potentially a Ph.D., with the ultimate goal of becoming a professor in the future. 👨🏼‍🏫

Selected Papers

An Overview of the Discrete Logarithm Problem in Cryptography
A Sarkar, D Guha Roy, P Datta
International Conference on Advanced Computing and Applications 2024

AgroGenie: A Smart Approach to Agriculture Using Machine Learning
A Sarkar, K Ganguly, S Ghosh, D Ghosh, S Saha, P Datta, DG Roy
International Conference on Information Science and Applications 2024

Exploring IoT for real-time CO2 monitoring and analysis
A Sarkar, D Ghosh, K Ganguly, S Ghosh, S Saha
arXiv preprint

Developed an AI-driven chatbot using Google’s Gemini api key and Langchain framework to provide detailed bird-related information.
Integrated Wikipedia API and custom bird encyclopedia PDFs for accurate and diverse data sources using RAG (Retrieval-augmented generation).
Code @ Github

First research experience, including the conception and implementation of a review paper.
Learned to read, understand, and draw conclusions from research papers.
Gained insights into progress in public-key cryptography research.
Presented the first draft as a Poster.
Accepted and Presented in ICACA-2024.

This research involves the utilization of a Machine-learning-based approach to analyze the values of several parameters influencing crop growth. It then accepts a novel set of input parameters and recommends the most probable crops that can be cultivated.
Adequate preprocessing methods have been applied to ensure the optimum removal of partiality from data. The ML models have been compared for their performance using a well-accepted validation measure, with Random Forest achieving a 99.6% accuracy.
Then, crops with similar features have been clustered to provide more flexible farming options for the farmer.
Accepted and Presented in ICISA-2024.

Worked in a team of 6 designing an IoT-enabled CO2 monitoring system using components such as NodeMCU-esp8266, mq135(gas sensor), & dht-11(temperature & humidity Sensor). Used ThingSpeak to plot data from the device.
Provides real-time CO2 level monitoring in ppm, and alarms the user through a buzzer and LED light when the CO2 level exceeds 500 ppm.
I worked on writing and debugging the code in Arduino IDE and calibrating the mq135.
Link to ArXiv Article