The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. ex. Some numerals are expressed as "XNUMX".
Copyrights notice
The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. Copyrights notice
산업계에서는 자동 이상 탐지에 대한 수요가 높습니다. 이상 현상은 거의 또는 전혀 나타나지 않는 샘플 그룹으로 정의되었습니다. 제품 유형에 따라 수많은 샘플을 수집하고 이상 탐지기를 훈련해야 합니다. 재고가 충분한 기존 유형의 제품으로 학습된 모델을 새로운 유형으로 전환하면 생산 라인이 구축되기 전에 새로운 유형의 이상을 감지할 수 있습니다. 그러나 이상현상의 정의로 인해 일반적인 이상현상 탐지기는 새로운 유형의 제품이 표준과 일치하더라도 이상현상으로 간주한다. 위의 실제 요구를 고려하여 본 연구에서는 소스 도메인에서 훈련된 이상 탐지기가 전체 재훈련 없이 소규모 대상 샘플 세트에 적용되는 새로운 문제 설정인 소수 샷 이상 탐지를 제안합니다. 그런 다음 딥러닝을 기반으로 한 계층적 확률 모델을 사용하여 이 문제를 해결합니다. 장난감 및 실제 데이터 세트에 대한 경험적 결과는 제안된 모델이 작은 대상 샘플 세트에서 이상 현상을 성공적으로 감지한다는 것을 보여줍니다.
Kazuki SATO
Kobe University
Satoshi NAKATA
Mitsubishi Chemical Systems, Inc.
Takashi MATSUBARA
Osaka University
Kuniaki UEHARA
Osaka Gakuin University
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
부
Kazuki SATO, Satoshi NAKATA, Takashi MATSUBARA, Kuniaki UEHARA, "Few-Shot Anomaly Detection Using Deep Generative Models for Grouped Data" in IEICE TRANSACTIONS on Information,
vol. E105-D, no. 2, pp. 436-440, February 2022, doi: 10.1587/transinf.2021EDL8063.
Abstract: There exists a great demand for automatic anomaly detection in industrial world. The anomaly has been defined as a group of samples that rarely or never appears. Given a type of products, one has to collect numerous samples and train an anomaly detector. When one diverts a model trained with old types of products with sufficient inventory to the new type, one can detect anomalies of the new type before a production line is established. However, because of the definition of the anomaly, a typical anomaly detector considers the new type of products anomalous even if it is consistent with the standard. Given the above practical demand, this study propose a novel problem setting, few-shot anomaly detection, where an anomaly detector trained in source domains is adapted to a small set of target samples without full retraining. Then, we tackle this problem using a hierarchical probabilistic model based on deep learning. Our empirical results on toy and real-world datasets demonstrate that the proposed model detects anomalies in a small set of target samples successfully.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.2021EDL8063/_p
부
@ARTICLE{e105-d_2_436,
author={Kazuki SATO, Satoshi NAKATA, Takashi MATSUBARA, Kuniaki UEHARA, },
journal={IEICE TRANSACTIONS on Information},
title={Few-Shot Anomaly Detection Using Deep Generative Models for Grouped Data},
year={2022},
volume={E105-D},
number={2},
pages={436-440},
abstract={There exists a great demand for automatic anomaly detection in industrial world. The anomaly has been defined as a group of samples that rarely or never appears. Given a type of products, one has to collect numerous samples and train an anomaly detector. When one diverts a model trained with old types of products with sufficient inventory to the new type, one can detect anomalies of the new type before a production line is established. However, because of the definition of the anomaly, a typical anomaly detector considers the new type of products anomalous even if it is consistent with the standard. Given the above practical demand, this study propose a novel problem setting, few-shot anomaly detection, where an anomaly detector trained in source domains is adapted to a small set of target samples without full retraining. Then, we tackle this problem using a hierarchical probabilistic model based on deep learning. Our empirical results on toy and real-world datasets demonstrate that the proposed model detects anomalies in a small set of target samples successfully.},
keywords={},
doi={10.1587/transinf.2021EDL8063},
ISSN={1745-1361},
month={February},}
부
TY - JOUR
TI - Few-Shot Anomaly Detection Using Deep Generative Models for Grouped Data
T2 - IEICE TRANSACTIONS on Information
SP - 436
EP - 440
AU - Kazuki SATO
AU - Satoshi NAKATA
AU - Takashi MATSUBARA
AU - Kuniaki UEHARA
PY - 2022
DO - 10.1587/transinf.2021EDL8063
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E105-D
IS - 2
JA - IEICE TRANSACTIONS on Information
Y1 - February 2022
AB - There exists a great demand for automatic anomaly detection in industrial world. The anomaly has been defined as a group of samples that rarely or never appears. Given a type of products, one has to collect numerous samples and train an anomaly detector. When one diverts a model trained with old types of products with sufficient inventory to the new type, one can detect anomalies of the new type before a production line is established. However, because of the definition of the anomaly, a typical anomaly detector considers the new type of products anomalous even if it is consistent with the standard. Given the above practical demand, this study propose a novel problem setting, few-shot anomaly detection, where an anomaly detector trained in source domains is adapted to a small set of target samples without full retraining. Then, we tackle this problem using a hierarchical probabilistic model based on deep learning. Our empirical results on toy and real-world datasets demonstrate that the proposed model detects anomalies in a small set of target samples successfully.
ER -