video classification using svm

I extracted features from some short video files and saved each of them in separate 3d matrices. Now, how can I use SVM as classifier?

[Walter Roberson]

svm would not care what the origin of the data was, so you should be able to use regular svm techniques.

[shab shekan]

Could you please explain regular svm techniques?

what's your suggestion?

Answer 1

[Kenta]

One option is to use CNN and LSTM.

The demo below is for video classification using machine/deep leaning method. It depends, but I think you can get better result with LSTM (Long Short Term Memory).

https://jp.mathworks.com/matlabcentral/fileexchange/74402-video-classification-using-lstm-lstm

The image features were extracted via a pre-trained network and the time-series features were classified using LSTM (Long Short Term Memory).

Answer 2

[Thijs]

if you have the bioinformatic toolbox you can use svmtrain to train your classifier and svmclassify to classify it.

[shab shekan]

thank you. I knew that commands and i used them before. but now, svmclassify give me no answer in plot.

so i think i didn't set svm properly with this data structure that explained before.