decode

Class: Autoencoder

(To be removed) Decode encoded data

decode will be removed in a future release. For more information, see Transition Legacy Neural Network Code to dlnetwork Workflows.

For advice on updating your code, see Version History.

Syntax

Y = decode(autoenc,Z)

Description

Y = decode(autoenc,Z)returns the decoded data Y, using the autoencoder autoenc.

Input Arguments

autoenc — Trained autoencoder

Autoencoder object

Trained autoencoder, returned by the trainAutoencoder function as an object of the Autoencoder class.

Z — Data encoded by autoenc

matrix

Data encoded by autoenc, specified as a matrix. Each column of Z represents an encoded sample (observation).

Data Types: single | double

Output Arguments

Y — Decoded data

matrix | cell array of image data

Decoded data, returned as a matrix or a cell array of image data.

If the autoencoder autoenc was trained on a cell array of image data, then Y is also a cell array of images.

If the autoencoder autoenc was trained on a matrix, then Y is also a matrix, where each column of Y corresponds to one sample or observation.

Examples

Decode Encoded Data For New Images

Load the training data.

X = digitTrainCellArrayData;

X is a 1-by-5000 cell array, where each cell contains a 28-by-28 matrix representing a synthetic image of a handwritten digit.

Train an autoencoder using the training data with a hidden size of 15.

hiddenSize = 15;
autoenc = trainAutoencoder(X,hiddenSize);

Extract the encoded data for new images using the autoencoder.

Xnew = digitTestCellArrayData;
features = encode(autoenc,Xnew);

Decode the encoded data from the autoencoder.

Y = decode(autoenc,features);

Y is a 1-by-5000 cell array, where each cell contains a 28-by-28 matrix representing a synthetic image of a handwritten digit.

Algorithms

If the input to an autoencoder is a vector $$x\in {ℝ}^{D_x}$$, then the encoder maps the vector x to another vector $$z\in {ℝ}^{D^{\left(1\right)}}$$ as follows:

$$z=h^{{}^{\left(1\right)}}\left(W^{\left(1\right)}x+b^{\left(1\right)}\right),$$

where the superscript (1) indicates the first layer. $$h^{\left(1\right)}:{ℝ}^{D^{\left(1\right)}}\to {ℝ}^{D^{\left(1\right)}}$$ is a transfer function for the encoder, $$W^{\left(1\right)}\in {ℝ}^{D^{\left(1\right)}\times D_{{}^x}}$$ is a weight matrix, and $$b^{\left(1\right)}\in {ℝ}^{D^{\left(1\right)}}$$ is a bias vector. Then, the decoder maps the encoded representation z back into an estimate of the original input vector, x, as follows:

$$\widehat{x}=h^{{}^{\left(2\right)}}\left(W^{\left(2\right)}z+b^{\left(2\right)}\right),$$

where the superscript (2) represents the second layer. $$h^{\left(2\right)}:{ℝ}^{D_x}\to {ℝ}^{D_x}$$ is the transfer function for the decoder,$$W^{\left(1\right)}\in {ℝ}^{D_{{}^x}\times D^{\left(1\right)}}$$ is a weight matrix, and $$b^{\left(2\right)}\in {ℝ}^{D_x}$$ is a bias vector.

Version History

Introduced in R2015b

R2026a: decode will be removed

decode will be removed in a future release.

To learn more about autoencoder workflows, see these examples and topics:

• Build Networks with Deep Network Designer

• Time Series Anomaly Detection Using Deep Learning

• Detect Anomalies Using Wavelet Scattering with Autoencoders

• Train Variational Autoencoder (VAE) to Generate Images

• List of Deep Learning Layer Blocks and Subsystems

• Code Generation for Deep Learning Networks

For more information, see Transition Legacy Neural Network Code to dlnetwork Workflows.