search

jbdel / OMG_UMONS_submission

5 stars 5 forks source link

add modules of classification and F1 score for emotion classes #1

Open yongxuUSTC opened 6 years ago

yongxuUSTC commented 6 years ago

This is very interesting project.

Could you also add classification modules for 7 emotion classes and calculate the F1 score?

I am new for emotion recognition. I just wonder why the OMG challenge did not give the emotion classification, the valence and arousal are not so distinct for me.

jbdel commented 6 years ago

Hello,

The challenge was to predict arousal and valence, but they also gave the emotion (see column EmotionMaxVote here : https://github.com/knowledgetechnologyuhh/OMGEmotionChallenge/blob/master/omg_TrainVideos.csv) going from 0 to 6 : 0 - Anger 1 - Disgust 2 - Fear 3 - Happy 4 - Neutral 5 - Sad 6 - Surprise

I wont modify or maintain this project because of a lack of time, but thank you for your interest.

The code should be rather easy to modify though.

Best,

yongxuUSTC commented 6 years ago

Hello,

Do you have the reference labels for the test set? Thanks

Best regards, yong

Yong XU

From: jbdel Date: 2018-08-22 01:28 To: jbdel/OMG_UMONS_submission CC: yong xu @ seattle; Author Subject: Re: [jbdel/OMG_UMONS_submission] add modules of classification and F1 score for emotion classes (#1) Hello, The challenge was to predict arousal and valence, but they also gave the emotion (see column EmotionMaxVote here : https://github.com/knowledgetechnologyuhh/OMGEmotionChallenge/blob/master/omg_TrainVideos.csv) going from 0 to 6 : 0 - Anger 1 - Disgust 2 - Fear 3 - Happy 4 - Neutral 5 - Sad 6 - Surprise I wont modify or maintain this project because of a lack of time, but thank you for your interest. The code should be rather easy to modify though. Best, — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or mute the thread.