Naman Birla

Software Developer at DP World | ML and AI Enthusiast | IIITD’26 | B.Tech CSE

Delhi, India

About

I’m a Computer Science undergraduate at IIIT Delhi with a strong passion for Machine Learning, Deep Learning, and Computer Vision. Currently, I’m conducting research at the Signal Processing Lab, where I’m working on applying deep learning algorithms to improve underwater image segmentation and enhancement. I’ve gained hands-on experience through several impactful projects. I contributed to a Foreign Object Debris (FOD) Detection system in collaboration with the Indian Air Force, where I developed and optimized CNN and YOLO-based models, achieving near-perfect accuracy. I also worked on a Brain Tumor Classification model using a medical imaging dataset, leveraging classical ML and deep learning approaches to achieve high accuracy. Recently, I’ve been exploring the theoretical and practical aspects of Complex-Valued Neural Networks, diving deeper into their potential applications in vision tasks and signal processing. I am always excited to connect and collaborate on AI and ML projects!

Experience

  • Software Engineer at DP World
    Jul 2026 - Present · 1 mo

  • Undergraduate Student Researcher at SBILab
    Dec 2024 - Dec 2025 · 1 yr 1 mo

  • AI/ML Developer at Mployee.me
    May 2025 - Jun 2025 · 2 mos

  • Undergraduate Teaching Assistant at Indraprastha Institute of Information Technology, Delhi
    Jan 2025 - May 2025 · 5 mos

    Undergraduate teaching assistant for Statistical Machine Learning (CSE342) course at Indraprastha Institute of Information Technology, Delhi.

  • Undergraduate Student Researcher at Indraprastha Institute of Information Technology, Delhi
    May 2024 - Jul 2024 · 3 mos

    Worked as an Undergraduate Researcher at the Visual Conception Group, IIIT Delhi, where I focused on developing an advanced solution for Foreign Object Debris (FOD) detection on airplane runways in collaboration with the Indian Air Force. Leveraging deep learning techniques, I built and annotated a custom dataset of over 4000 frames extracted from runway videos. I implemented and optimized ResNet18 and YOLO models using PyTorch, achieving near 100% accuracy in detecting FODs. My work involved data preprocessing, model training, and integrating these models to enhance detection accuracy and minimize false positives. This experience honed my skills in computer vision, neural networks, and collaborative research.