RoBIC: A benchmark suite for assessing classifiers robustness
Many defenses have emerged with the development of adversarial attacks. Models must be objectively evaluated accordingly. This paper systematically tackles this concern by proposing a new parameter-free benchmark we coin RoBIC. RoBIC fairly evaluates the robustness of image classifiers using a new half-distortion measure. It gauges the robustness of the network against white and black box attacks, independently of its accuracy. RoBIC is faster than the other available benchmarks. We present the significant differences in the robustness of 16 recent models as assessed by RoBIC.
Code (1)
Tasks
Adversarial AttackSimilar Papers 제목 키워드 기반
Electrodermal Activity as a Unimodal Signal for Aerobic Exercise Detection in Wearable Sensors
Electrodermal Activity (EDA) is a non-invasive physiological signal widely available in wearable devices and reflects sympathetic nervous system (SNS) activation. Prior multi-modal studies have demonstrated robust perfor…
Assessing Performance of Aerobic Routines using Background Subtraction and Intersected Image Region
It is recommended for a novice to engage a trained and experience person, i.e., a coach before starting an unfamiliar aerobic or weight routine. The coach's task is to provide real-time feedbacks to ensure that the routi…
The effects of temperature acclimation on swimming performance in the pelagic Mahi-mahi (Coryphaena hippurus)
Mahi-mahi (Coryphaena hippurus) are a highly migratory pelagic fish, but little is known about what environmental factors drive their broad distribution. This study examined how temperature influences aerobic scope and s…
Statistical Multicriteria Benchmarking via the GSD-Front
Given the vast number of classifiers that have been (and continue to be) proposed, reliable methods for comparing them are becoming increasingly important. The desire for reliability is broken down into three main aspect…
BenchmarkingBayesian Uncertainty Quantification for Anaerobic Digestion models
Uncertainty quantification is critical for ensuring adequate predictive power of computational models used in biology. Focusing on two anaerobic digestion models, this article introduces a novel generalized Bayesian proc…
Inductive BiasUncertainty Quantification