深度学习在信号处理领域的应用

使用信号处理应用扩展深度学习工作流

通过将 Deep Learning Toolbox™ 与 Signal Processing Toolbox™ 或 Wavelet Toolbox™ 结合使用，将深度学习应用于信号处理。有关音频和语音处理领域的应用，请参阅深度学习在音频处理领域的应用。有关无线通信中的应用，请参阅使用深度学习的无线通信。

App

信号标注器

Label signal attributes, regions, and points of interest, and extract features

函数

`labeledSignalSet`	Create labeled signal set
`signalLabelDefinition`	Create signal label definition
`signalMask`	Modify and convert signal masks and extract signal regions of interest
`countlabels`	Count number of unique labels
`folders2labels`	Get list of labels from folder names
`splitlabels`	Find indices to split labels according to specified proportions
`signalDatastore`	Datastore for collection of signals
`dlstft`	Deep learning short-time Fourier transform
`stftLayer`	Short-time Fourier transform layer

主题

Pedestrian and Bicyclist Classification Using Deep Learning (Radar Toolbox)
Classify pedestrians and bicyclists based on their micro-Doppler characteristics using a deep learning network and time-frequency analysis.
Radar and Communications Waveform Classification Using Deep Learning (Radar Toolbox)
Classify radar and communications waveforms using the Wigner-Ville distribution (WVD) and a deep convolutional neural network (CNN).
Automate Signal Labeling with Custom Functions (Signal Processing Toolbox)
Use Signal Labeler to locate and label QRS complexes and R peaks of ECG signals.
Crack Identification from Accelerometer Data (Wavelet Toolbox)
Use wavelet and deep learning techniques to detect transverse pavement cracks and localize their position.
Deploy Signal Segmentation Deep Network on Raspberry Pi
Generate a MEX function and a standalone executable to perform waveform segmentation on a Raspberry Pi™.
Deploy Signal Classifier on NVIDIA Jetson Using Wavelet Analysis and Deep Learning
This example shows how to generate and deploy a CUDA® executable that classifies human electrocardiogram (ECG) signals using features extracted by the continuous wavelet transform (CWT) and a pretrained convolutional neural network (CNN).
Deploy Signal Classifier Using Wavelets and Deep Learning on Raspberry Pi
This example shows the workflow to classify human electrocardiogram (ECG) signals using the Continuous Wavelet Transform (CWT) and a deep convolutional neural network (CNN).

特色示例

使用深度学习进行波形分割

此示例说明如何使用递归深度学习网络和时频分析来分割人体心电图 (ECG) 信号。

打开实时脚本

使用长短期记忆网络对 ECG 信号进行分类

此示例说明如何使用深度学习和信号处理对来自 PhysioNet 2017 Challenge 的心电图 (ECG) 数据进行分类。具体而言，该示例使用长短期记忆网络和时频分析。

打开实时脚本

使用小波分析和深度学习对时间序列分类

此示例说明如何使用连续小波变换 (CWT) 和深度卷积神经网络 (CNN) 对人体心电图 (ECG) 信号进行分类。

打开实时脚本

Generate Synthetic Signals Using Conditional GAN

Use a conditional generative adversarial network to produce synthetic data for model training.

打开实时脚本

Learn Pre-Emphasis Filter Using Deep Learning

Use a convolutional deep network to learn a pre-emphasis filter for speech recognition.

打开实时脚本

Denoise EEG Signals Using Deep Learning Regression

Remove EOG noise from EEG signals using deep learning regression.

Denoise Signals with Adversarial Learning Denoiser Model

Use an adversarial learning denoiser model to denoise signals.

Hand Gesture Classification Using Radar Signals and Deep Learning

Classify ultra-wideband impulse radar signal data using a MISO convolutional neural network.

打开实时脚本

Parasite Classification Using Wavelet Scattering and Deep Learning

Classify parasitic infections using wavelet image scattering and deep learning.

打开实时脚本

Fault Detection Using Wavelet Scattering and Recurrent Deep Networks

Classify faults in acoustic recordings of air compressors using a wavelet scattering network paired with a recurrent neural network.

打开实时脚本

基于小波散射和深度学习的口述数字识别

此示例说明如何使用机器学习和深度学习方法对语音数字进行分类。在此示例中，您将使用小波时间散射配合支持向量机 (SVM) 和长短期记忆 (LSTM) 网络执行分类。您还可以应用贝叶斯优化来确定合适的超参数，以提高 LSTM 网络的准确度。此外，该示例说明了一种使用深度卷积神经网络 (CNN) 和 Mel 频谱图的方法。

打开实时脚本