Outlier-aware Inlier Modeling and Multi-scale Scoring for Anomalous Sound Detection via Multitask Learning

被引:1
|
作者
Zhang, Yucong [1 ]
Suo, Hongbin [2 ]
Wan, Yulong [2 ]
Li, Ming [1 ]
机构
[1] Duke Kunshan Univ, Data Sci Res Ctr, Kunshan, Peoples R China
[2] OPPO, Data & AI Engn Syst, Beijing, Peoples R China
来源
基金
中国国家自然科学基金;
关键词
anomalous sound detection; multitask learning; outlier exposure; inlier modeling;
D O I
10.21437/Interspeech.2023-572
中图分类号
O42 [声学];
学科分类号
070206 ; 082403 ;
摘要
This paper proposes an approach for anomalous sound detection that incorporates outlier exposure and inlier modeling within a unified framework by multitask learning. While outlier exposure-based methods can extract features efficiently, it is not robust. Inlier modeling is good at generating robust features, but the features are not very effective. Recently, serial approaches are proposed to combine these two methods, but it still requires a separate training step for normal data modeling. To overcome these limitations, we use multitask learning to train a conformer-based encoder for outlier-aware inlier modeling. Moreover, our approach provides multi-scale scores for detecting anomalies. Experimental results on the MIMII and DCASE 2020 task 2 datasets show that our approach outperforms state-of-the-art single-model systems and achieves comparable results with top-ranked multi-system ensembles.
引用
收藏
页码:5381 / 5385
页数:5
相关论文
共 21 条
  • [1] GAMSNet: Globally aware road detection network with multi-scale residual learning
    Lu, Xiaoyan
    Zhong, Yanfei
    Zheng, Zhuo
    Zhang, Liangpei
    ISPRS JOURNAL OF PHOTOGRAMMETRY AND REMOTE SENSING, 2021, 175 : 340 - 352
  • [2] Multi-scale Context-aware User Interest Learning for Behavior Pattern Modeling
    Deng, Zhiying
    Li, Jianjun
    Zou, Li
    Liu, Wei
    Shi, Si
    Chen, Qian
    Zhao, Juan
    Li, Guohui
    DATABASE SYSTEMS FOR ADVANCED APPLICATIONS, DASFAA 2024, PT 3, 2025, 14852 : 333 - 342
  • [3] Multi-Exposure Image Fusion via Multi-Scale and Context-Aware Feature Learning
    Liu, Yu
    Yang, Zhigang
    Cheng, Juan
    Chen, Xun
    IEEE SIGNAL PROCESSING LETTERS, 2023, 30 : 100 - 104
  • [4] MSCAF-Net: A General Framework for Camouflaged Object Detection via Learning Multi-Scale Context-Aware Features
    Liu, Yu
    Li, Haihang
    Cheng, Juan
    Chen, Xun
    IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY, 2023, 33 (09) : 4934 - 4947
  • [5] Normality Learning-based Graph Anomaly Detection via Multi-Scale Contrastive Learning
    Duan, Jingcan
    Zhang, Pei
    Wang, Siwei
    Hu, Jingtao
    Jin, Hu
    Zhang, Jiaxin
    Zhou, Haifang
    Liu, Xinwang
    PROCEEDINGS OF THE 31ST ACM INTERNATIONAL CONFERENCE ON MULTIMEDIA, MM 2023, 2023, : 7502 - 7511
  • [6] MODELING FUNCTIONAL NETWORK DYNAMICS VIA MULTI-SCALE DICTIONARY LEARNING AND NETWORK CONTINUUMS
    Li, Xiang
    Lin, Binbin
    Lv, Jinglei
    Ye, Jieping
    Liu, Tianming
    2016 IEEE 13TH INTERNATIONAL SYMPOSIUM ON BIOMEDICAL IMAGING (ISBI), 2016, : 66 - 69
  • [7] IMAGE SALIENCY DETECTION VIA MULTI-SCALE STATISTICAL NON-REDUNDANCY MODELING
    Scharfenberger, Christian
    Jain, Aanchal
    Wong, Alexander
    Fieguth, Paul
    2014 IEEE INTERNATIONAL CONFERENCE ON IMAGE PROCESSING (ICIP), 2014, : 4294 - 4298
  • [8] Graph Anomaly Detection via Multi-Scale Contrastive Learning Networks with Augmented View
    Duan, Jingcan
    Wang, Siwei
    Zhang, Pei
    Zhu, En
    Hu, Jingtao
    Jin, Hu
    Liu, Yue
    Dong, Zhibin
    THIRTY-SEVENTH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE, VOL 37 NO 6, 2023, : 7459 - 7467
  • [9] Crowd-Level Abnormal Behavior Detection via Multi-Scale Motion Consistency Learning
    Luo, Linbo
    Li, Yuanjing
    Yin, Haiyan
    Xie, Shangwei
    Hu, Ruimin
    Cai, Wentong
    THIRTY-SEVENTH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE, VOL 37 NO 7, 2023, : 8984 - 8992
  • [10] MSFFAL: Few-Shot Object Detection via Multi-Scale Feature Fusion and Attentive Learning
    Zhang, Tianzhao
    Sun, Ruoxi
    Wan, Yong
    Zhang, Fuping
    Wei, Jianming
    SENSORS, 2023, 23 (07)