Self-Tuning Self-Supervised Image Anomaly Detection

TL;DR

This paper introduces ST-SSAD, an unsupervised method for tuning data augmentation in self-supervised image anomaly detection, leading to improved accuracy without labeled validation data.

Contribution

The work presents the first end-to-end unsupervised augmentation tuning approach for SSAD, including a new validation loss and differentiable augmentation functions.

Findings

Significant performance improvements over existing methods.

Effective augmentation tuning without labeled validation data.

Validated on industrial defect and semantic anomaly datasets.

Abstract

Self-supervised learning (SSL) has emerged as a promising paradigm that presents supervisory signals to real-world problems, bypassing the extensive cost of manual labeling. Consequently, self-supervised anomaly detection (SSAD) has seen a recent surge of interest, since SSL is especially attractive for unsupervised tasks. However, recent works have reported that the choice of a data augmentation function has significant impact on the accuracy of SSAD, posing augmentation search as an essential but nontrivial problem due to lack of labeled validation data. In this paper, we introduce ST-SSAD, the first unsupervised approach to end-to-end augmentation tuning for SSAD. To this end, our work presents two key contributions. The first is a new unsupervised validation loss that quantifies the alignment between augmented training data and unlabeled validation data. The second is new differentiable augmentation functions, allowing data augmentation hyperparameter(s) to be tuned in an end-to-end manner. Experiments on two testbeds with semantic class anomalies and subtle industrial defects show that ST-SSAD gives significant performance gains over existing works. All our code and testbeds are available at https://github.com/jaeminyoo/ST-SSAD.