A Weak Supervision Approach to Detecting Visual Anomalies for Automated Testing of Graphics Units

TL;DR

This paper introduces a deep learning system that automatically detects visual corruptions in graphics units using weak supervision, eliminating the need for manual labeling and outperforming existing unsupervised methods.

Contribution

The work presents a novel weak supervision approach leveraging driver bugs and MIL techniques for real-time visual anomaly detection in graphics units.

Findings

Outperforms unsupervised GAN-based models

Detects novel visual corruptions

Operates efficiently in real-time

Abstract

We present a deep learning system for testing graphics units by detecting novel visual corruptions in videos. Unlike previous work in which manual tagging was required to collect labeled training data, our weak supervision method is fully automatic and needs no human labelling. This is achieved by reproducing driver bugs that increase the probability of generating corruptions, and by making use of ideas and methods from the Multiple Instance Learning (MIL) setting. In our experiments, we significantly outperform unsupervised methods such as GAN-based models and discover novel corruptions undetected by baselines, while adhering to strict requirements on accuracy and efficiency of our real-time system.