Abstract:The rapid progress in machine learning models has significantly boosted the potential for real-world applications such as autonomous vehicles, disease diagnoses, and recognition of emergencies. The performance of many machine learning models depends on the nature and size of the training data sets. These models often face challenges due to the scarcity, noise, and imbalance in real-world data, limiting their performance. Nonetheless, high-quality, diverse, relevant and representative training data is essential to build accurate and reliable machine learning models that adapt well to real-world scenarios. It is hypothesised that well-designed synthetic data can improve the performance of a machine learning algorithm. This work aims to create a synthetic dataset and evaluate its effectiveness to improve the prediction accuracy of object detection systems. This work considers autonomous vehicle scenarios as an illustrative example to show the efficacy of synthetic data. The effectiveness of these synthetic datasets in improving the performance of state-of-the-art object detection models is explored. The findings demonstrate that incorporating synthetic data improves model performance across all performance matrices. Two deep learning systems, System-1 (trained on real-world data) and System-2 (trained on a combination of real and synthetic data), are evaluated using the state-of-the-art YOLO model across multiple metrics, including accuracy, precision, recall, and mean average precision. Experimental results revealed that System-2 outperformed System-1, showing a 3% improvement in accuracy, along with superior performance in all other metrics.

Instance-Level Safety-Aware Fidelity of Synthetic Data and Its Calibration

Observability-Aware Intrinsic and Extrinsic Calibration of LiDAR-IMU Systems

Fill In The Gaps: Model Calibration and Generalization with Synthetic Data

An Adaptive Multi-Fidelity Sampling Framework for Safety Analysis of Connected and Automated Vehicles

How Safe Am I Given What I See? Calibrated Prediction of Safety Chances for Image-Controlled Autonomy

Enhancing Object Detection Accuracy in Autonomous Vehicles Using Synthetic Data

Hands-Up: Leveraging Synthetic Data for Hands-On-Wheel Detection

Safety-Enhanced Autonomous Driving Using Interpretable Sensor Fusion Transformer

Adaptive Fusion of Single-View and Multi-View Depth for Autonomous Driving

Safety-aware Causal Representation for Trustworthy Offline Reinforcement Learning in Autonomous Driving

Online LiDAR-Camera Extrinsic Parameters Self-checking and Recalibration

On the Equivalency, Substitutability, and Flexibility of Synthetic Data

Synthetic training data generation for deep learning based quality inspection

Joint Falsification and Fidelity Settings Optimization for Validation of Safety-Critical Systems: A Theoretical Analysis

Harnessing Machine Learning for Discerning AI-Generated Synthetic Images

What Makes and Breaks Safety Fine-tuning? A Mechanistic Study

Good Data Is All Imitation Learning Needs

CalQNet -- Detection of Calibration Quality for Life-Long Stereo Camera Setups

Improving Uncertainty Calibration of Deep Neural Networks via Truth Discovery and Geometric Optimization