Instance-Level Safety-Aware Fidelity of Synthetic Data and Its Calibration
This work addresses safety-critical applications like self-driving technology by enhancing synthetic data fidelity, though it appears incremental as it builds on existing methods for data generation and calibration.
The paper tackles the problem of ensuring synthetic data accurately reflects real-world safety issues in self-driving applications by introducing four types of instance-level fidelity and an optimization method to refine synthetic data generators. Experiments show this tuning improves the correlation between safety-critical errors in synthetic and real data.
Modeling and calibrating the fidelity of synthetic data is paramount in shaping the future of safe and reliable self-driving technology by offering a cost-effective and scalable alternative to real-world data collection. We focus on its role in safety-critical applications, introducing four types of instance-level fidelity that go beyond mere visual input characteristics. The aim is to ensure that applying testing on synthetic data can reveal real-world safety issues, and the absence of safety-critical issues when testing under synthetic data can provide a strong safety guarantee in real-world behavior. We suggest an optimization method to refine the synthetic data generator, reducing fidelity gaps identified by deep learning components. Experiments show this tuning enhances the correlation between safety-critical errors in synthetic and real data.