Get Global Guarantees: On the Probabilistic Nature of Perturbation Robustness
This addresses robustness assessment for safety-critical deep learning applications, though it appears incremental as it builds on existing definitions and methodologies.
The paper tackles the problem of efficiently assessing neural network robustness against imperceptible perturbations in safety-critical applications, proposing a novel metric called tower robustness based on hypothesis testing that enables more rigorous and efficient pre-deployment evaluations.
In safety-critical deep learning applications, robustness measures the ability of neural models that handle imperceptible perturbations in input data, which may lead to potential safety hazards. Existing pre-deployment robustness assessment methods typically suffer from significant trade-offs between computational cost and measurement precision, limiting their practical utility. To address these limitations, this paper conducts a comprehensive comparative analysis of existing robustness definitions and associated assessment methodologies. We propose tower robustness to evaluate robustness, which is a novel, practical metric based on hypothesis testing to quantitatively evaluate probabilistic robustness, enabling more rigorous and efficient pre-deployment assessments. Our extensive comparative evaluation illustrates the advantages and applicability of our proposed approach, thereby advancing the systematic understanding and enhancement of model robustness in safety-critical deep learning applications.