Performance validation of Multi-Object Tracking Algorithm with Radar Sensor

This project focuses on enhancing a tool designed to validate and evaluate multi-object tracking (MOT) algorithms used with 77GHz radar sensors. Currently, the tool faces some challenges:

1. It struggles with processing large or complex datasets efficiently, leading to delays.
2. The tool’s evaluation metrics are still limited; the research paper should be more implemented in order to evaluate radar’s performance more efficient and effective.
3. In certain scenarios, especially under cluttered environments or high-speed motion, the tool may misidentify or lose track of objects.

To address these challenges, we aim to:

• Optimize processing time by reducing it by approximately 25%, improving the tool’s efficiency when handling large datasets.
• Enhance visualization output, fixing video time sync (no delay between frames)
• Research and integrate advanced evaluation metrics tailored to radar field, ensuring a more accurate analysis applying in different real scenarios
• Support for specific configurations from OEMs, providing a more robust, flexible and scalable solution for tracking validation. OEMs have their own mounting position and use cases for radar setups, support for this enable a more versatile system/tool to handle specific client needs