LADMIM: Logical Anomaly Detection with Masked Image Modeling in Discrete Latent Space
This addresses the challenge of detecting anomalies in feature relationships for industrial applications, representing an incremental improvement over existing methods.
The paper tackles the problem of detecting logical anomalies in industrial images, such as incorrect object combinations or positional deviations, by proposing a novel approach that uses masked image modeling in a discrete latent space, achieving an average AUC of 0.867 on the MVTecLOCO dataset.
Detecting anomalies such as incorrect combinations of objects or deviations in their positions is a challenging problem in industrial anomaly detection. Traditional methods mainly focus on local features of normal images, such as scratches and dirt, making detecting anomalies in the relationships between features difficult. Masked image modeling(MIM) is a self-supervised learning technique that predicts the feature representation of masked regions in an image. To reconstruct the masked regions, it is necessary to understand how the image is composed, allowing the learning of relationships between features within the image. We propose a novel approach that leverages the characteristics of MIM to detect logical anomalies effectively. To address blurriness in the reconstructed image, we replace pixel prediction with predicting the probability distribution of discrete latent variables of the masked regions using a tokenizer. We evaluated the proposed method on the MVTecLOCO dataset, achieving an average AUC of 0.867, surpassing traditional reconstruction-based and distillation-based methods.