DDR-ID: Dual Deep Reconstruction Networks Based Image Decomposition for Anomaly Detection
This work addresses anomaly detection in images, which is crucial for applications like security and quality control, but it appears incremental as it builds on existing reconstruction-based methods with specific enhancements.
The paper tackles the challenge of learning discriminative information from normal-class images for anomaly detection by proposing DDR-ID, a method that decomposes images into normal and residual components using dual deep reconstruction networks, and it outperforms benchmarking methods on MNIST, CIFAR-10, Endosome, and GTSRB datasets.
One pivot challenge for image anomaly (AD) detection is to learn discriminative information only from normal class training images. Most image reconstruction based AD methods rely on the discriminative capability of reconstruction error. This is heuristic as image reconstruction is unsupervised without incorporating normal-class-specific information. In this paper, we propose an AD method called dual deep reconstruction networks based image decomposition (DDR-ID). The networks are trained by jointly optimizing for three losses: the one-class loss, the latent space constrain loss and the reconstruction loss. After training, DDR-ID can decompose an unseen image into its normal class and the residual components, respectively. Two anomaly scores are calculated to quantify the anomalous degree of the image in either normal class latent space or reconstruction image space. Thereby, anomaly detection can be performed via thresholding the anomaly score. The experiments demonstrate that DDR-ID outperforms multiple related benchmarking methods in image anomaly detection using MNIST, CIFAR-10 and Endosome datasets and adversarial attack detection using GTSRB dataset.