Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

A Real-Time DETR Approach to Bangladesh Road Object Detection for Autonomous Vehicles

Irfan Nafiz Shahan, Arban Hossain, Saadman Sakib, Al-Mubin Nabil

TL;DR

Real-Time DETR (RTDETR) object detection on the BadODD Road Object Detection dataset based in Bangladesh is used, and real-time DETR models are shown to perform significantly better on inference times, with minimal loss of accuracy and performance.

Abstract

In the recent years, we have witnessed a paradigm shift in the field of Computer Vision, with the forthcoming of the transformer architecture. Detection Transformers has become a state of the art solution to object detection and is a potential candidate for Road Object Detection in Autonomous Vehicles. Despite the abundance of object detection schemes, real-time DETR models are shown to perform significantly better on inference times, with minimal loss of accuracy and performance. In our work, we used Real-Time DETR (RTDETR) object detection on the BadODD Road Object Detection dataset based in Bangladesh, and performed necessary experimentation and testing. Our results gave a mAP50 score of 0.41518 in the public 60% test set, and 0.28194 in the private 40% test set.

A Real-Time DETR Approach to Bangladesh Road Object Detection for Autonomous Vehicles

TL;DR

Real-Time DETR (RTDETR) object detection on the BadODD Road Object Detection dataset based in Bangladesh is used, and real-time DETR models are shown to perform significantly better on inference times, with minimal loss of accuracy and performance.

Abstract

In the recent years, we have witnessed a paradigm shift in the field of Computer Vision, with the forthcoming of the transformer architecture. Detection Transformers has become a state of the art solution to object detection and is a potential candidate for Road Object Detection in Autonomous Vehicles. Despite the abundance of object detection schemes, real-time DETR models are shown to perform significantly better on inference times, with minimal loss of accuracy and performance. In our work, we used Real-Time DETR (RTDETR) object detection on the BadODD Road Object Detection dataset based in Bangladesh, and performed necessary experimentation and testing. Our results gave a mAP50 score of 0.41518 in the public 60% test set, and 0.28194 in the private 40% test set.

Paper Structure

This paper contains 9 sections, 5 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Methodology
  3. Understanding BadODD
  4. Model Selection
  5. Data Preprocessing
  6. Model Configuration and Hyperparameter Tuning
  7. Training
  8. Results and Discussion
  9. Conclusion and Future Work

Figures (5)

  • Figure 1: Dataset Analysis of BadODD, showcasing the disparity in model class occurrences and the uneven distribution of bounding box sizes,
  • Figure 5: Ground Truth Labels showing congested persons
  • Figure 6: Predicted Labels Trying to Identify the congested labels
  • Figure 7: Performance Metrics for the RTDETR-X model
  • Figure 8: Confusion Matrix for each of the categories