File Exchange

image thumbnail

Video classification using LSTM(LSTMによる動画の分類)

version 1.3.2 (58.9 MB) by Kenta
simple example of video classification with LSTM

9 Downloads

Updated 16 Mar 2020

GitHub view license on GitHub

[English]
This is a simple example of video classification using LSTM with MATLAB.
Please run the code named VideoClassificationExample.
This example was created based on a Mathworks official documentation located at
https://jp.mathworks.com/help/deeplearning/examples/classify-videos-using-deep-learning.html
While the official example requires down-loading a dataset about 2 GB, this example can try that
with a small amout of data, which may help you giving a try easily.
Note that this is just an example of LSTM with images and please refer to the official example for your further study.
I appreciate for the free pictures from used in the thumbnail and live editor (https://www.irasutoya.com/).
[Japanese]
深層学習を用いてビデオの分類を行います。その人が歩いているのか/走っているのかをその人の頭に取り付けたカメラの動画から予測します。動画のフレームを入力とし、学習済みネットワークにより特徴量を取り出します。そして、その特徴量からLSTMによる分類を行います。静止画の分類は多く紹介されていますが、ビデオを入力とし、その数秒間のビデオから対象が何であるかを分類する例はmatlab document中にあまり多くありませんでした。また公式ドキュメントにも例はありますが、2ギガのデータセットをダウンロードする必要があり、ダウンロードや計算に多くの時間がかかり、手軽に試すにはやや不向きです。参考になれば幸いです。

[References]
[1] Matlab Official Documentation: Classify Videos Using Deep Learning
https://jp.mathworks.com/help/deeplearning/ug/classify-videos-using-deep-learning.html
[2] Irasutoya: https://www.irasutoya.com/ (images in the script were obtained from this website)

Cite As

Kenta (2020). Video classification using LSTM(LSTMによる動画の分類) (https://github.com/giants19/video_classification_LSTM_matlab), GitHub. Retrieved .

Comments and Ratings (4)

Shunichi Kusano

Yoshito Saito

michio

Takuji Fukumoto

Updates

1.3.2

Description added

1.3

Description added

1.2

Description added

1.1

other requirements added

1.0.11

Licence file added

1.0.10

Acknowledgement added

1.0.9

Acknowledgement added

1.0.8

Acknowledgement added

1.0.7

.md file was added

1.0.6

Japanese subtitle added

1.0.5

readme changed

1.0.4

image changed

1.0.3

image changed

1.0.2

image changed

1.0.1

title change

MATLAB Release Compatibility
Created with R2019b
Compatible with any release
Platform Compatibility
Windows macOS Linux