GIT: Detecting Uncertainty, Out-Of-Distribution and Adversarial Samples using Gradients and Invariance Transformations
This addresses safety-critical applications by providing a holistic detection method for generalization errors, though it appears incremental as it builds on existing detection approaches.
The paper tackled the problem of detecting generalization errors in deep neural networks, such as uncertainty, out-of-distribution, and adversarial samples, by proposing GIT, which combines gradient information and invariance transformations, and demonstrated superior performance compared to state-of-the-art methods across various architectures and setups.
Deep neural networks tend to make overconfident predictions and often require additional detectors for misclassifications, particularly for safety-critical applications. Existing detection methods usually only focus on adversarial attacks or out-of-distribution samples as reasons for false predictions. However, generalization errors occur due to diverse reasons often related to poorly learning relevant invariances. We therefore propose GIT, a holistic approach for the detection of generalization errors that combines the usage of gradient information and invariance transformations. The invariance transformations are designed to shift misclassified samples back into the generalization area of the neural network, while the gradient information measures the contradiction between the initial prediction and the corresponding inherent computations of the neural network using the transformed sample. Our experiments demonstrate the superior performance of GIT compared to the state-of-the-art on a variety of network architectures, problem setups and perturbation types.