International Conference on Computer and Knowledge Engineering

Home / 14th International Conference on Computer and Knowledge Engineering

Bridging the Synthetic-to-Real Gap (BSRG): Creating Simulated Datasets for Domain Adaptation to Enhance Vehicle Detection

Authors :

Behnaz Sadeghigol¹ Mohammad Ali Keyvanrad²

1- Faculty of Electrical & Computer Engineering Malek Ashtar University of Technology 2- Faculty of Electrical & Computer Engineering Malek Ashtar University of Technology

Keywords :

Synthetic dataset،Unreal Engine،Object detection،Domain adaptation،Transfer learning،JLTV dataset

Abstract :

Deep neural network based military vehicle detectors pose particular challenges due to the scarcity of relevant images and limited access to vehicles in this domain. Moreover, Real-world data often poses significant challenges, including privacy, availability, and bias. To mitigate these challenges, synthetic datasets can be leveraged. This article explores the efficacy of synthetic datasets in training state-of-the-art object detection models, specifically focusing on the Joint Light Tactical Vehicle (JLTV). Using the powerful Unreal Engine, which can create highly realistic scenes, we generated a comprehensive synthetic dataset designed to simulate real-world conditions and enhance the training process for various detection algorithms. In this study, we evaluate two distinct models for object recognition: an enhanced domain matching approach utilizing the Masked Image Consistency (MIC) framework and an unsupervised domain matching approach employing confidence-based mixing (ConfMix). The MIC model achieved a mean Average Precision (mAP) at 50% of 47% on real-world data, while the ConfMix model attained a mAP@50 of 55%. These results underscore the pivotal role of synthetic data in advancing object recognition technologies. They also highlight potential research directions for improving synthetic dataset generation and enhancing model performance in practical applications. Examples of this dataset can be accessed at: https://github.com/behnaz-sadeghigol/JLTV_dataset.