Debjit Mandal

Welcome to My GitHub Page. Type 'help' to see available commands.

Debjit Mandal - Resume

Education

Kalinga Institute of Industrial Technology, Bhubaneswar, Odisha
B.Tech, Computer Science and Engineering (2022 - 2026)
CGPA: 9.13/10

Experience

Campus Ambassador, IMUNA (July 2023 - August 2023)

Promoted IMUN Online Conferences in different schools/universities and got 10+ registrations.
Awarded Best Intern for a week.

Projects

NeuroSeg: A Deep Learning-based Brain Tumor Segmentation System.
- Designed and trained a UNet-based deep learning model for automatic brain tumor segmentation from MRI images.
- Developed a full-stack web application using Flask (backend) and HTML/CSS/JS (frontend) for real-time tumor detection.
ECGShield: An interactive Streamlit-based system for ECG noise removal.
- Implemented various signal filtering techniques including Notch Filter, High-Pass, Low-Pass, and Wavelet Denoising.
- Designed a user-friendly Streamlit interface to upload ECG files, visualize signals, and download the cleaned output.
MoodMelody: A Real-time emotion detection and content recommendation system.
- Implemented emotion recognition using TensorFlow/Keras.
- Integrated Spotify API for personalized music recommendations.
DSH: A Command-Line shell made using C++.
- Implemented custom command parsing for user-defined commands.
- Handled concurrency with multithreading for running background processes.
TrafficVisionAI: Real-Time Traffic Sign Recognition and Driver Assistance using AI.
- Integrated traffic sign recognition using TensorFlow and OpenCV.
- Developed lane detection and collision warning systems.

Patents

UK Design Patent – Autonomous Vehicle Predictive Steering Device
Design No. 6424920 (Registered on 20 February 2025; Granted on 07 March 2025)
Co-held with MR. PRITAM CHAKRABORTY, MR. NILOTPAL BASU, and DR. ANJAN BANDYOPADHYAY.

Skills

Programming Languages: Python, C/C++
Web Technologies: HTML, CSS
Tools: Git
Frameworks & Libraries: TensorFlow, OpenCV, scikit-learn, Pandas, NumPy
Databases: SQLPlus