Data Shift of Object Detection in Autonomous Driving

TL;DR

This paper investigates data shift issues in autonomous driving object detection, analyzing its complexity, reviewing detection methods, and proposing an optimized YOLOv5 model with CycleGAN augmentation that outperforms baselines on BDD100K.

Contribution

It systematically analyzes data shift in autonomous driving, reviews detection methods, and introduces an improved YOLOv5 model with CycleGAN augmentation for better robustness.

Findings

Our method outperforms baseline models on BDD100K.

CycleGAN augmentation improves detection robustness.

Data shift significantly impacts autonomous driving detection performance.

Abstract

With the widespread adoption of machine learning technologies in autonomous driving systems, their role in addressing complex environmental perception challenges has become increasingly crucial. However, existing machine learning models exhibit significant vulnerability, as their performance critically depends on the fundamental assumption that training and testing data satisfy the independent and identically distributed condition, which is difficult to guarantee in real-world applications. Dynamic variations in data distribution caused by seasonal changes, weather fluctuations lead to data shift problems in autonomous driving systems. This study investigates the data shift problem in autonomous driving object detection tasks, systematically analyzing its complexity and diverse manifestations. We conduct a comprehensive review of data shift detection methods and employ shift detection analysis techniques to perform dataset categorization and balancing. Building upon this foundation, we construct an object detection model. To validate our approach, we optimize the model by integrating CycleGAN-based data augmentation techniques with the YOLOv5 framework. Experimental results demonstrate that our method achieves superior performance compared to baseline models on the BDD100K dataset.