An Ensemble Stacked Convolutional Neural Network Model for Environmental Event Sound Recognition

Li, Shaobo; Yao, Yong; Hu, Jie; Liu, Guokai; Yao, Xuemei; Hu, Jianjun

doi:10.3390/app8071152

Cited by 92 publications

(63 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this study, through a series of classifications for three study sites, we found that: (1) The proposed CNN-Polygon approach works well for land cover classification even when the number of training samples is very small, and (2) the proposed CNN-Matrix performs well when multi-temporal data are used for classification and the training sample size is relatively large. In particular, this study showed that the types (and structures) of input features are a critical consideration of CNN-based classification [73,74]. More visually intuitive input features tend to increase the classification accuracy even when training data are limited.…”

Section: Novelty and Limitationsmentioning

confidence: 84%

Different Spectral Domain Transformation for Land Cover Classification Using Convolutional Neural Networks with Multi-Temporal Satellite Imagery

et al. 2020

View full text Add to dashboard Cite

This study compares some different types of spectral domain transformations for convolutional neural network (CNN)-based land cover classification. A novel approach was proposed, which transforms one-dimensional (1-D) spectral vectors into two-dimensional (2-D) features: Polygon graph images (CNN-Polygon) and 2-D matrices (CNN-Matrix). The motivations of this study are that (1) the shape of the converted 2-D images is more intuitive for human eyes to interpret when compared to 1-D spectral input; and (2) CNNs are highly specialized and may be able to similarly utilize this information for land cover classification. Four seasonal Landsat 8 images over three study areas—Lake Tapps, Washington, Concord, New Hampshire, USA, and Gwangju, Korea—were used to evaluate the proposed approach for nine land cover classes compared to several other methods: Random forest (RF), support vector machine (SVM), 1-D CNN, and patch-based CNN. Oversampling and undersampling approaches were conducted to examine the effect of the sample size on the model performance. The CNN-Polygon had better performance than the other methods, with overall accuracies of about 93%–95 % for both Concord and Lake Tapps and 80%–84% for Gwangju. The CNN-Polygon particularly performed well when the training sample size was small, less than 200 per class, while the CNN-Matrix resulted in similar or higher performance as sample sizes became larger. The contributing input variables to the models were carefully analyzed through sensitivity analysis based on occlusion maps and accuracy decreases. Our result showed that a more visually intuitive representation of input features for CNN-based classification models yielded higher performance, especially when the training sample size was small. This implies that the proposed graph-based CNNs would be useful for land cover classification where reference data are limited.

show abstract

Section: Novelty and Limitationsmentioning

confidence: 84%

Different Spectral Domain Transformation for Land Cover Classification Using Convolutional Neural Networks with Multi-Temporal Satellite Imagery

et al. 2020

View full text Add to dashboard Cite

show abstract

“…As shown on Table II, for the ESC-10 dataset, TimeScaleNet only allows to achieve environmental sound classification with a mean accuracy of 69.71% and a standard deviation of 1.91% across the five folds. This result is far from matching the best results on environmental sound classification using raw audio on the ESC-10 dataset [53]. In [53], the authors described RawNet, whose intent is also to achieve joint feature learning in the time domain, along with sound classification.…”

Section: B Environmental Sound Classification Performance Evaluationmentioning

confidence: 91%

“…This result is far from matching the best results on environmental sound classification using raw audio on the ESC-10 dataset [53]. In [53], the authors described RawNet, whose intent is also to achieve joint feature learning in the time domain, along with sound classification. Their approach allowed to achieve 85.2% of accuracy, which is much better than the obtained performance of TimeScaleNet using the ESC-10 dataset, which only slightly outperforms the baseline methods proposed by the maintainer of the dataset in [48] and [54].…”

Section: B Environmental Sound Classification Performance Evaluationmentioning

confidence: 91%

“…We also suspect that the rather moderate performances of TimeScaleNet for ESC could be linked to the fact that the number of parameters of TimeScaleNet are too large for such a small sized dataset. As a comparison, the number of learnt parameters used by Li et al in [53] is 1.14 M, which is approximately 10 times smaller than in TimeScaleNet, for the same ESC task.…”

Section: B Environmental Sound Classification Performance Evaluationmentioning

confidence: 99%

See 1 more Smart Citation

TimeScaleNet: A Multiresolution Approach for Raw Audio Recognition Using Learnable Biquadratic IIR Filters and Residual Networks of Depthwise-Separable One-Dimensional Atrous Convolutions

Bavu

Ramamonjy

Pujol

et al. 2019

IEEE J. Sel. Top. Signal Process.

View full text Add to dashboard Cite

In the present paper, we show the benefit of a multi-resolution approach that allows to encode the relevant information contained in unprocessed time domain acoustic signals. TimeScaleNet aims at learning an efficient representation of a sound, by learning time dependencies both at the sample level and at the frame level. The proposed approach allows to improve the interpretability of the learning scheme, by unifying advanced deep learning and signal processing techniques. In particular, TimeScaleNet's architecture introduces a new form of recurrent neural layer, which is directly inspired from digital IIR signal processing. This layer acts as a learnable passband biquadratic digital IIR filterbank. The learnable filterbank allows to build a time-frequency-like feature map that self-adapts to the specific recognition task and dataset, with a large receptive field and very few learnable parameters. The obtained frame-level feature map is then processed using a residual network of depthwise separable atrous convolutions. This second scale of analysis aims at efficiently encoding relationships between the time fluctuations at the frame timescale, in different learnt pooled frequency bands, in the range of [20 ms ; 200 ms]. TimeScaleNet is tested both using the Speech Commands Dataset and the ESC-10 Dataset. We report a very high mean accuracy of 94.87 ± 0.24% (macro averaged F1-score : 94.9 ± 0.24%) for speech recognition, and a rather moderate accuracy of 69.71 ± 1.91% (macro averaged F1score : 70.14 ± 1.57%) for the environmental sound classification task.

show abstract

“…Recently, deep learning has been shown to achieve impressive results in areas such as computer vision, image and video processing, speech, and natural language processing [19]. Deep learning methods have also been applied to fault diagnosis [20][21][22] and the RUL estimation problem. For the RUL estimation problem, a Convolutional Neural Network (CNN) is applied on a sliding window with multi-times weight and failures height in [23].…”

Section: Introductionmentioning

confidence: 99%

Transfer Learning with Deep Recurrent Neural Networks for Remaining Useful Life Estimation

et al. 2018

Self Cite

View full text Add to dashboard Cite

Prognostics, such as remaining useful life (RUL) prediction, is a crucial task in condition-based maintenance. A major challenge in data-driven prognostics is the difficulty of obtaining a sufficient number of samples of failure progression. However, for traditional machine learning methods and deep neural networks, enough training data is a prerequisite to train good prediction models. In this work, we proposed a transfer learning algorithm based on Bi-directional Long Short-Term Memory (BLSTM) recurrent neural networks for RUL estimation, in which the models can be first trained on different but related datasets and then fine-tuned by the target dataset. Extensive experimental results show that transfer learning can in general improve the prediction models on the dataset with a small number of samples. There is one exception that when transferring from multi-type operating conditions to single operating conditions, transfer learning led to a worse result.

show abstract

An Ensemble Stacked Convolutional Neural Network Model for Environmental Event Sound Recognition

Cited by 92 publications

References 22 publications

Different Spectral Domain Transformation for Land Cover Classification Using Convolutional Neural Networks with Multi-Temporal Satellite Imagery

Different Spectral Domain Transformation for Land Cover Classification Using Convolutional Neural Networks with Multi-Temporal Satellite Imagery

TimeScaleNet: A Multiresolution Approach for Raw Audio Recognition Using Learnable Biquadratic IIR Filters and Residual Networks of Depthwise-Separable One-Dimensional Atrous Convolutions

Transfer Learning with Deep Recurrent Neural Networks for Remaining Useful Life Estimation

Contact Info

Product

Resources

About