Sound Event Detection with Depthwise Separable and Dilated Convolutions

Drossos, Konstantinos; Mimilakis, Stylianos Ioannis; Gharib, Shayan; Li, Yanxiong; Virtanen, Tuomas

doi:10.1109/ijcnn48605.2020.9207532

Cited by 32 publications

(60 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…In a paper previously published on audio classification, CNN was replaced by depthwise separable (DWS) convolution [ 29 , 30 ]. DWS was a decomposition form of standard convolution, which decomposed a standard convolution into one convolution and one 1 × 1 convolution (called pointwise convolution) [ 31 ].…”

Section: Related Workmentioning

confidence: 99%

Underwater Acoustic Target Recognition Based on Depthwise Separable Convolution Neural Networks

Wang

Liu

2021

Sensors

View full text Add to dashboard Cite

Facing the complex marine environment, it is extremely challenging to conduct underwater acoustic target feature extraction and recognition using ship-radiated noise. In this paper, firstly, taking the one-dimensional time-domain raw signal of the ship as the input of the model, a new deep neural network model for underwater target recognition is proposed. Depthwise separable convolution and time-dilated convolution are used for passive underwater acoustic target recognition for the first time. The proposed model realizes automatic feature extraction from the raw data of ship radiated noise and temporal attention in the process of underwater target recognition. Secondly, the measured data are used to evaluate the model, and cluster analysis and visualization analysis are performed based on the features extracted from the model. The results show that the features extracted from the model have good characteristics of intra-class aggregation and inter-class separation. Furthermore, the cross-folding model is used to verify that there is no overfitting in the model, which improves the generalization ability of the model. Finally, the model is compared with traditional underwater acoustic target recognition, and its accuracy is significantly improved by 6.8%.

show abstract

Section: Related Workmentioning

confidence: 99%

Underwater Acoustic Target Recognition Based on Depthwise Separable Convolution Neural Networks

Wang

Liu

2021

Sensors

View full text Add to dashboard Cite

show abstract

“…To have an idea of the effect of depthwise convolutions, it is possible to compare one of the early CNNs such as AlexNet [22] with YAMNet. The number of parameters contained in AlexNet amounts to 61 million in 8 layers, whereas YAMNet contains 3.7 million parameters [41] in 30 layers. Fewer learnable weights further reduce overfitting risks in such architectures.…”

Section: Yamnet: An Efficient Cnn For Sound Event Detectionmentioning

confidence: 99%

“…However, no previous studies have investigated knowledge transfer from sound recognition tasks to machine diagnosis. The present work is motivated by the idea that networks pre-trained on audios for sound event detection (SED) [40][41][42] may encapsulate the necessary knowledge to classify REB spectrograms. The main goal of SED is to identify instances of sound events in audio recordings [40,43].…”

Section: Introductionmentioning

confidence: 99%

Deep Transfer Learning for Machine Diagnosis: From Sound and Music Recognition to Bearing Fault Detection

2021

View full text Add to dashboard Cite

Today’s deep learning strategies require ever-increasing computational efforts and demand for very large amounts of labelled data. Providing such expensive resources for machine diagnosis is highly challenging. Transfer learning recently emerged as a valuable approach to address these issues. Thus, the knowledge learned by deep architectures in different scenarios can be reused for the purpose of machine diagnosis, minimizing data collecting efforts. Existing research provides evidence that networks pre-trained for image recognition can classify machine vibrations in the time-frequency domain by means of transfer learning. So far, however, there has been little discussion about the potentials included in networks pre-trained for sound recognition, which are inherently suited for time-frequency tasks. This work argues that deep architectures trained for music recognition and sound detection can perform machine diagnosis. The YAMNet convolutional network was designed to serve extremely efficient mobile applications for sound detection, and it was originally trained on millions of data extracted from YouTube clips. That framework is employed to detect bearing faults for the CWRU dataset. It is shown that transferring knowledge from sound and music recognition to bearing fault detection is successful. The maximum accuracy is achieved using a few hundred data for fine-tuning the fault diagnosis model.

show abstract

“…The aforementioned observation was proven by several researchers [3][4][5][6][7][8][9][10][11][12][13][14]. DL is beneficial in other fields, including target recognition [15], speech recognition [16,17], image recognition [18][19][20], image restoration [21][22][23], audio classification [24,25], object detection [26][27][28][29][30], scene recognition [31], etc., but it has been considered "bad news" in text-based CAPTCHAs, by penetrating their security and making them vulnerable.…”

Section: Introductionmentioning

confidence: 98%

Securing IoT Devices: A Robust and Efficient Deep Learning with a Mixed Batch Adversarial Generation Process for CAPTCHA Security Verification

Dankwa

Yang

2021

Electronics

View full text Add to dashboard Cite

The Internet of Things environment (e.g., smart phones, smart televisions, and smart watches) ensures that the end user experience is easy, by connecting lives on web services via the internet. Integrating Internet of Things devices poses ethical risks related to data security, privacy, reliability and management, data mining, and knowledge exchange. An adversarial machine learning attack is a good practice to adopt, to strengthen the security of text-based CAPTCHA (Completely Automated Public Turing test to tell Computers and Humans Apart), to withstand against malicious attacks from computer hackers, to protect Internet of Things devices and the end user’s privacy. The goal of this current study is to perform security vulnerability verification on adversarial text-based CAPTCHA, based on attacker–defender scenarios. Therefore, this study proposed computation-efficient deep learning with a mixed batch adversarial generation process model, which attempted to break the transferability attack, and mitigate the problem of catastrophic forgetting in the context of adversarial attack defense. After performing K-fold cross-validation, experimental results showed that the proposed defense model achieved mean accuracies in the range of 82–84% among three gradient-based adversarial attack datasets.

show abstract

Sound Event Detection with Depthwise Separable and Dilated Convolutions

Cited by 32 publications

References 20 publications

Underwater Acoustic Target Recognition Based on Depthwise Separable Convolution Neural Networks

Underwater Acoustic Target Recognition Based on Depthwise Separable Convolution Neural Networks

Deep Transfer Learning for Machine Diagnosis: From Sound and Music Recognition to Bearing Fault Detection

Securing IoT Devices: A Robust and Efficient Deep Learning with a Mixed Batch Adversarial Generation Process for CAPTCHA Security Verification

Contact Info

Product

Resources

About