Anomaly detection using Variational Autoencoder(VAE)

You can learn how to detect and localize anomalies on image using Variational Autoencoder
916 Downloads
Updated 25 Dec 2020

On shipping inspection for chemical materials, clothing, and food materials, etc, it is necessary to detect defects and impurities in normal products.
In the following link, I shared codes to detect and localize anomalies using CAE with only images for training.

In this demo, you can learn how to apply Variational Autoencoder(VAE) to this task instead of CAE.
VAEs use a probability distribution on the latent space, and sample from this distribution to generate new data.

[Japanese]
正常な画像のみ使ってCAEモデルを学習させ,正常な画像に紛れる異常をディープラーニングを用いて検出ならびに位置の特定を行えるコードを下記のリンクで紹介しました。
このデモでは代わりにVariational Autoencoderを適用した
方法をご紹介します。
VAEは潜在変数に確率分布を使用し、この分布からサンプリングして新しいデータを生成するものです。

■Anomaly detection and localization using deep learning(CAE)
https://jp.mathworks.com/matlabcentral/fileexchange/72444-anomaly-detection-and-localization-using-deep-learning-cae

[Keyward] 画像処理・ディープラーニング・DeepLearning・IPCVデモ ・異常検出・外観検査・オートエンコーダー・サンプルコード・変分オートエンコーダ

■Auto-Encoding Variational Bayes [2013]
Diederik P Kingma, Max Welling
https://arxiv.org/pdf/1312.6114.pdf

Cite As

Takuji Fukumoto (2024). Anomaly detection using Variational Autoencoder(VAE) (https://github.com/mathworks/Anomaly-detection-using-Variational-Autoencoder-VAE-/releases/tag/1.0.1), GitHub. Retrieved .

MATLAB Release Compatibility
Created with R2019b
Compatible with R2019b and later releases
Platform Compatibility
Windows macOS Linux
Categories
Find more on AI for Audio in Help Center and MATLAB Answers

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!
Version Published Release Notes
1.0.1

See release notes for this release on GitHub: https://github.com/mathworks/Anomaly-detection-using-Variational-Autoencoder-VAE-/releases/tag/1.0.1

1.0.0

To view or report issues in this GitHub add-on, visit the GitHub Repository.
To view or report issues in this GitHub add-on, visit the GitHub Repository.