Improving Reconstruction Autoencoder Out-of-distribution Detection with Mahalanobis Distance

Abstract

There is an increasingly apparent need for validating the classifications made by deep learning systems in safety-critical applications like autonomous vehicle systems. A number of recent papers have proposed methods for detecting anomalous image data that appear different from known inlier data samples, including reconstruction-based autoencoders. Autoencoders optimize the compression of input data to a latent space of a dimensionality smaller than the original input and attempt to accurately reconstruct the input using that compressed representation. Since the latent vector is optimized to capture the salient features from the inlier class only, it is commonly assumed that images of objects from outside of the training class cannot effectively be compressed and reconstructed. Some thus consider reconstruction error as a kind of novelty measure. Here we suggest that reconstruction-based approaches fail to capture particular anomalies that lie far from known inlier samples in latent space but near the latent dimension manifold defined by the parameters of the model. We propose incorporating the Mahalanobis distance in latent space to better capture these out-of-distribution samples and our results show that this method often improves performance over the baseline approach.

9 pages, 5 figures

References in corpus (1)

• A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networks

Cited by in corpus (14)

• Likelihood Regret: An Out-of-Distribution Detection Score For Variational Auto-encoder
• Why is the Mahalanobis Distance Effective for Anomaly Detection?
• Detecting Out-of-Distribution Inputs in Deep Neural Networks Using an Early-Layer Output
• MTS-DVGAN: Anomaly Detection in Cyber-Physical Systems using a Dual Variational Generative Adversarial Network
• Diffusion Policies for Out-of-Distribution Generalization in Offline Reinforcement Learning
• FRODO: Free rejection of out-of-distribution samples: application to chest x-ray analysis
• Analysis of Confident-Classifiers for Out-of-distribution Detection
• Anomalous Example Detection in Deep Learning: A Survey
• Detecting Out-of-distribution Samples via Variational Auto-encoder with Reliable Uncertainty Estimation
• Do We Really Need to Learn Representations from In-domain Data for Outlier Detection?
• Resisting Out-of-Distribution Data Problem in Perturbation of XAI
• Class Anchor Clustering: a Loss for Distance-based Open Set Recognition
• Conformance Constraint Discovery: Measuring Trust in Data-Driven Systems
• Dynamic Horizon Value Estimation for Model-based Reinforcement Learning