TY - JOUR
T1 - Calibrated reconstruction based adversarial autoencoder model for novelty detection
AU - Huang, Yi
AU - Li, Ying
AU - Jourjon, Guillaume
AU - Seneviratne, Suranga
AU - Thilakarathna, Kanchana
AU - Cheng, Adriel
AU - Webb, Darren
AU - Xu, Richard Yi Da
N1 - This research was supported by an Australian Government Research Training Program Scholarship and conducted in partnership with the Defence Science and Technology Group and Data61-CSIRO, through the Next Generation Technologies Fund.
Publisher Copyright:
© 2023 Elsevier B.V.
PY - 2023/5
Y1 - 2023/5
N2 - Novelty detection detects outliers located at any location, such as abnormalities (i.e., far distance outliers) and novel/unobserved patterns (i.e., close distance outliers). While many novelty detection approaches have been proposed in the literature, they generally focus on detecting one specific type of outlier, e.g., Multi-Class Open Set Recognition (MCOSR) and One-Class Novelty Detection (OCND) approaches are applied for far and close distance outlier detection, respectively. However, in practice, it is difficult to measure in advance whether the distance between outliers and inliers is far or close. Recent work on outlier detection at any location with a unified model has yielded mixed performance. In this paper, we propose a new unified model, named Calibrated Reconstruction Based Adversarial AutoEncoder (CRAAE), for location agnostic outlier detection. The key idea is to integrate implicit and explicit confidence calibration strategies into a reconstruction based model for building a more accurate decision boundary. We leverage the category information disentangled from feature space to calibrate the decision metric (i.e., reconstruction error) constructed in the original data space. CRAAE also adds Uniform or Dirichlet noise into the artificial outlier generation process to represent various outliers. Experimental results show that CRAAE can outperform state-of-the-art unified models (e.g., GPND) and achieve similar performance with OCND and MCOSR methods in close and far distance outlier detection, respectively.
AB - Novelty detection detects outliers located at any location, such as abnormalities (i.e., far distance outliers) and novel/unobserved patterns (i.e., close distance outliers). While many novelty detection approaches have been proposed in the literature, they generally focus on detecting one specific type of outlier, e.g., Multi-Class Open Set Recognition (MCOSR) and One-Class Novelty Detection (OCND) approaches are applied for far and close distance outlier detection, respectively. However, in practice, it is difficult to measure in advance whether the distance between outliers and inliers is far or close. Recent work on outlier detection at any location with a unified model has yielded mixed performance. In this paper, we propose a new unified model, named Calibrated Reconstruction Based Adversarial AutoEncoder (CRAAE), for location agnostic outlier detection. The key idea is to integrate implicit and explicit confidence calibration strategies into a reconstruction based model for building a more accurate decision boundary. We leverage the category information disentangled from feature space to calibrate the decision metric (i.e., reconstruction error) constructed in the original data space. CRAAE also adds Uniform or Dirichlet noise into the artificial outlier generation process to represent various outliers. Experimental results show that CRAAE can outperform state-of-the-art unified models (e.g., GPND) and achieve similar performance with OCND and MCOSR methods in close and far distance outlier detection, respectively.
KW - Autoencoder
KW - Calibration
KW - Novelty detection
KW - Reconstruction
UR - http://www.scopus.com/inward/record.url?scp=85151647006&partnerID=8YFLogxK
U2 - 10.1016/j.patrec.2023.03.026
DO - 10.1016/j.patrec.2023.03.026
M3 - Journal article
AN - SCOPUS:85151647006
SN - 0167-8655
VL - 169
SP - 50
EP - 57
JO - Pattern Recognition Letters
JF - Pattern Recognition Letters
ER -