Micro Objet Detection
Implemented a micro object detection project for detecting small objects like buildings, planes, boats and vehicles in satellite imagery using YOLOv12s.
URL
Hello, I'm
Tushar Mitra
Final Year Computer Science Student | Data Science
By exploring varying areas of computing, I have developed a unique blend of skills. However, my academic journey has sparked a keen interest in Data Science and its related fields such as Machine Learning, Deep Learning, Big Data, Natural Language Processing, Computer Vision and Data Visualization tools like PowerBI and Tableau. These are the areas which I find both challenging and rewarding. This website is my platform for sharing the projects that I work on, and the technologies that I encounter in my explorations.
Education
Kalinga Institute of Industrial Technology
Bhubaneswar, Odisha
Bachelor of Technology
Major : Computer Science and Engineering
Minor : Financial Economics (with Data Analytics)
2022 - 2026
Loyola High School
Patna, Bihar
2009-2021
<skills>
Programming Languages
- C
- Python
- Java
Development
- HTML5
- CSS3
- MySQL
- MongoDB
Data Science & ML
PyTorch
Tensorflow
Scikit-Learn- Pandas
- Numpy
Matplotlib
Tools & Operating Systems
- Git
- Docker
- Linux
- Shell
Cloud Computing
- Amazon Web Services
- Google Cloud Platform
- Azure
<projects>
Talk to Pdf
Implemented a Generative AI project in a web app to upload PDFs and respond to queries related to those PDFs.
URL
Vision-based Attendance System
Implemented a Vision-based attendance system using computer vision and haarcascade frontalface classifier.
URL
ETL Pipeline
Implemented an ETL (Extract, Transform, Load) Pipeline using Apache Airflow, Astronomer, PostgreSQL and Open Meteo API.
URL
Credit Card Fraud Detection
Implemented a Regression ML algorithm to calculate the price of diamond based on 22 different input features like sex, education, marriage, age, balance due, amount paid, amount to pay, etc
URL
Diamond Price Prediction
Implemented a Regression ML algorithm to calculate the price of diamond based on 9 different input features like carat, cut, clarity, depth, etc.
URL<experience>
Aug 2024
to
Nov 2024
to
Nov 2024
Smart India Hackathon 2024
Internal Hackathon and Screening Round
- Our team "Tech Titans" Qualified for the Screening Round of "Smart India Hackathon". We came in the Top 45 teams out of the 297 registered teams from our University. We submitted our presentation and prototype for the Geolocation-based Attendance System under Problem Statement 1707 for GAIL(Gas Authority of India Limited), under the Ministry of Petroleum and Natural Gas.
- After the Screening Round, we got selected for the Internal Hackathon Round. In this round, our problem statement, presentation and prototype was given for submission to be graded by the internal team of SIH 2024.
<research>
Feb 2025
to
Mar 2025
to
Mar 2025
Food Choices and Sustainable Consumption Pattern
- This study investigates consumer perceptions and behaviours related to food choices and waste in India, integrating individual dietary habits with broader social awareness and potential policy solutions.
- A Survey of Indian Youth was done with a sample size of 97.
- Chosen as the best paper in Technical Session 2.4.
- Presented at 2nd National Conclave on Viksit Bharat @ 2047, Organized by KiiT School of Economics and Commerce. [Paper]
Nov 2024
to
April 2025
to
April 2025
Micro Object Detection
- Downloaded ’xView satellite imagery dataset’ and split it in 70:15:15 ratio for training, validation and testing images. Performed image annotations on them in V7 Labs Darwin by dividing the objects into 4 classes.
- Trained the model by taking a small subset of 60 images in the same split ratio using YOLOv12 small as a pre-trained model and fine-tuning its parameters. Achieved good mAP and other metrics for 3 of the 4 classes.
- To be Presented. [Paper]
<contact>
Whether you are interested in collaborating on a project, have a job opportunity in mind, or just want to connect and chat about tech, feel free to reach out. I would love to hear from you!