Provably Robust Detection of Out-of-distribution Data (almost) for free

Meinke, Alexander; Bitterwolf, Julian; Hein, Matthias

doi:10.48550/arxiv.2106.04260

Cited by 4 publications

(4 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In fact, many OOD detection systems are susceptible to adversarial attacks Sehwag et al [2019], Chen et al [2022]. Meinke et al [2021] show how to verify robustness to adversarial examples around given input samples. Berrada et al [2021] develop a general framework for certifying properties of output distributions of neural networks given constraints on the input distribution.…”

Section: Ood Detectionmentioning

confidence: 99%

Provably Bounding Neural Network Preimages

Kotha¹,

Brix²,

Kolter³

et al. 2023

Preprint

View full text Add to dashboard Cite

Most work on the formal verification of neural networks has focused on bounding forward images of neural networks, i.e., the set of outputs of a neural network that correspond to a given set of inputs (for example, bounded perturbations of a nominal input). However, many use cases of neural network verification require solving the inverse problem, i.e, over-approximating the set of inputs that lead to certain outputs. In this work, we present the first efficient bound propagation algorithm, INVPROP, for verifying properties over the preimage of a linearly constrained output set of a neural network, which can be combined with branchand-bound to achieve completeness. Our efficient algorithm allows multiple passes of intermediate bound refinements, which are crucial for tight inverse verification because the bounds of an intermediate layer depend on relaxations both before and after this layer. We demonstrate our algorithm on applications related to quantifying safe control regions for a dynamical system and detecting out-of-distribution inputs to a neural network. Our results show that in certain settings, we can find overapproximations that are over 2500× tighter than prior work while being 2.5× faster on the same hardware.

show abstract

Section: Ood Detectionmentioning

confidence: 99%

Provably Bounding Neural Network Preimages

Kotha¹,

Brix²,

Kolter³

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…As in OOD detection, OSR, AD and ND, being robust against adversarial attacks is crucial. Recent works in OSR [133], [134], ND [18], [135], and OOD detection [136], [137] have investigated the effects of adversarial attacks on models. However more is needed.…”

Section: Adversarial Robustnessmentioning

confidence: 99%

A Unified Survey on Anomaly, Novelty, Open-Set, and Out-of-Distribution Detection: Solutions and Future Challenges

Salehi¹,

Mirzaei²,

Hendrycks³

et al. 2021

Preprint

View full text Add to dashboard Cite

Machine learning models often encounter samples that are diverged from the training distribution. Failure to recognize an out-of-distribution (OOD) sample, and consequently assign that sample to an in-class label significantly compromises the reliability of a model. The problem has gained significant attention due to its importance for safety deploying models in open-world settings. Detecting OOD samples is challenging due to the intractability of modeling all possible unknown distributions. To date, several research domains tackle the problem of detecting unfamiliar samples, including anomaly detection, novelty detection, one-class learning, open set recognition, and out-of-distribution detection. Despite having similar and shared concepts, out-of-distribution, open-set, and anomaly detection have been investigated independently. Accordingly, these research avenues have not cross-pollinated, creating research barriers. While some surveys intend to provide an overview of these approaches, they seem to only focus on a specific domain without examining the relationship between different domains. This survey aims to provide a cross-domain and comprehensive review of numerous eminent works in respective areas while identifying their commonalities. Researchers can benefit from the overview of research advances in different fields and develop future methodology synergistically. Furthermore, to the best of our knowledge, while there are surveys in anomaly detection or one-class learning, there is no comprehensive or up-to-date survey on out-of-distribution detection, which our survey covers extensively. Finally, having a unified cross-domain perspective, we discuss and shed light on future lines of research, intending to bring these fields closer together.

show abstract

“…Much of the OOD Detection research literature available makes a number of strange assumptions; among these, the existence of a dataset consisting of Out-Of-Distribution samples to train a binary classifier [15]. To comprehensively cover what is OOD, this dataset would have to contain samples from the entirety of the space outside of the In-Distribution.…”

Section: Related Workmentioning

confidence: 99%

DOODLER: Determining Out-Of-Distribution Likelihood from Encoder Reconstructions

Kent,

2021

Preprint

View full text Add to dashboard Cite

Deep Learning models possess two key traits that, in combination, make their use in the real world a risky prospect. One, they do not typically generalize well outside of the distribution for which they were trained, and two, they tend to exhibit confident behavior regardless of whether or not they are producing meaningful outputs. While Deep Learning possesses immense power to solve realistic, high-dimensional problems, these traits in concert make it difficult to have confidence in their real-world applications. To overcome this difficulty, the task of Out-Of-Distribution (OOD) Detection has been defined, to determine when a model has received an input from outside of the distribution for which it is trained to operate. This paper introduces and examines a novel methodology, DOODLER, for OOD Detection, which directly leverages the traits which result in its necessity. By training a Variational Auto-Encoder (VAE) on the same data as another Deep Learning model, the VAE learns to accurately reconstruct In-Distribution (ID) inputs, but not to reconstruct OOD inputs, meaning that its failure state can be used to perform OOD Detection. Unlike other work in the area, DOODLER requires only very weak assumptions about the existence of an OOD dataset, allowing for more realistic application. DOODLER also enables pixel-wise segmentations of input images by OOD likelihood, and experimental results show that it matches or outperforms methodologies that operate under the same constraints.

show abstract

Provably Robust Detection of Out-of-distribution Data (almost) for free

Cited by 4 publications

References 9 publications

Provably Bounding Neural Network Preimages

Provably Bounding Neural Network Preimages

A Unified Survey on Anomaly, Novelty, Open-Set, and Out-of-Distribution Detection: Solutions and Future Challenges

DOODLER: Determining Out-Of-Distribution Likelihood from Encoder Reconstructions

Contact Info

Product

Resources

About