Data Science Intern
I am a Data Science Intern with experience in 2024 at Atsuya Technologies. I developed a YOLOv8-based real-time object detection system improving event logging efficiency. My current focus is on machine learning and AI model development.
Evaluated six deep learning models to classify six PV panel fault types, comparing sharpening, gamma correction and grayscale preprocessing on balanced vs. imbalanced data. Achieved 99.39% test accuracy with Swin Transformer-Tiny using gamma correction on the balanced dataset, the best-performing model overall.
Built a deep learning model using MobileNetV2 and deployed with Gradio for an interactive user experience.
Developed a fraud detection web app handling 1.2M+ imbalanced records with preprocessing, oversampling and Random Forest, deployed via Flask and Streamlit.
Implemented YOLO-based object detection with quadrant tracking and frame analysis, integrated with a PostgreSQL database via Flask API to store detection events with timestamps.
Analysed histopathology lung images using CNN and LBP feature extraction.
June 2024 – June 2024
Developed a credit card fraud detection web app, handling 1.2M+ imbalanced records with preprocessing and oversampling, achieving high accuracy using Random Forest and deploying via Flask and Streamlit for real-time predictions.
August 2024 – November 2024
Built a YOLOv8-based system for real-time object detection with quadrant tracking and frame analysis, using LabelImg/LabelMe for annotation. Designed and deployed a YOLOv8-based real-time object detection system with Flask API & PostgreSQL backend, improving event logging efficiency for streaming data. Integrated OCR and explored LLMs (Ollama) for text extraction and structuring.
M.Tech ,Artificial Intelligence •
B.Tech, Artificial Intelligence and Data Science • 2022 - 2026
8.25
HSC • 2022
88%
SSLC • 2020
82%
Secured 2nd Prize
Runner-Up
Interested in collaboration or just want to say hello? Feel free to reach out!