Sep 2024 – May 2025 · 8 mos · Kharagpur, West Bengal, India · On-site

• Object Detection:
• Utilized the YOLOv9 model for detecting objects in traffic scenes.
• Vehicle Tracking:
• Implemented ByteTrack and DeepSort algorithms to track traffic vehicles effectively.
• Trajectory Analysis:
• Conducted an analysis of vehicle trajectories using trajectory clustering techniques.
• Risk Identification:
• Extracted 3D information from videos to identify potential collision between vehicles.
• Integrated functionality to identify risky situations occurring in traffic, such as potential collisions or unsafe driving patterns.
• Application Development:
• Designed and developed an application for video uploads that performs:
• Real-time object detection and tracking using YOLOv9 and ByteTrack.
• Identifies risky traffic situations.
• Provides detailed statistics about the uploaded video, including metrics like:
• Relative vehicle speed
• Relative distance between vehicles
• Other relevant traffic insights.

Apr 2024 – Aug 2024 · 4 mos · Mandi, Himachal Pradesh, India · On-site

• Conducted mathematical modeling on EEG signal data using computational topology.
• Demonstrated chaoticity in signals using Largest Lyapunov Exponent and calculated appropriate time-delay using Mutual Information.
• Utilized deep learning models (EEGNet, ATCNet) to classify smell molecules, achieving twice the improvement in results.
• Employed GradCam Explainable AI to make the EEGNet model interpretable.
• Performed Granger Causality analysis to examine the predictive relationship between olfactory bulb signals and EEG activity.

Nov 2023 – Present · 2 yrs 7 mos

• Developed alphas on WorldQuant Brain Simulation platform that were back tested in historic data.