Acoustic Anomaly Detection for Machine Sounds based on Image Transfer Learning
Preprint
Abstract
In industrial applications, the early detection of malfunctioning factory machinery is crucial. In this paper, we consider acoustic malfunction detection via transfer learning. Contrary to the majority of current approaches which are based on deep autoencoders, we propose to extract features using neural networks that were pretrained on the task of image classification. We then use these features to train a variety of anomaly detection models and show that this improves results compared to convolutional autoencoders in recordings of four different factory machines in noisy environments. Moreover, we find that features extracted from ResNet based networks yield better results than those from AlexNet and Squeezenet. In our setting, Gaussian Mixture Models and One-Class Support Vector Machines achieve the best anomaly detection performance.Keywords
All Related Versions
- Version 1, 2020-06-05, ArXiv
- Version 2, 2020-12-11, ArXiv
- Published version: , 49.
This publication has 6 references indexed in Scilit:
- A Review of Deep Learning Based Methods for Acoustic Scene ClassificationApplied Sciences, 2020
- MIMII Dataset: Sound Dataset for Malfunctioning Industrial Machine Investigation and Inspection2019
- Snore Sound Classification Using Image-Based Deep Spectrum Features2017
- Deep Convolutional Neural Networks and Data Augmentation for Environmental Sound ClassificationIEEE Signal Processing Letters, 2017
- Very Deep Convolutional Neural Networks for Noise Robust Speech RecognitionIEEE/ACM Transactions on Audio, Speech, and Language Processing, 2016
- Deep Image Features in Music Information RetrievalInternational Journal of Electronics and Telecommunications, 2014