[mirtoolbox] Re: miremotion
- From: Olivier Lartillot <olartillot@xxxxxxxxx>
- To: mirtoolbox@xxxxxxxxxxxxx
- Date: Tue, 18 Jan 2011 15:17:33 +0200
Hi Mark, Diego and Alan.
Sorry for the late reply.
This is right, there does not seem to be any clear explanation in the
documentation yet, concerning the numerical scale used for the representation
of each emotion class or dimension. Basically, each class or dimension is
supposed to have values spanned in the interval [1,7]. Value 1 would correspond
to very low value, and value 7 to very high value. We use this convention
because we collected listeners' emotional ratings exactly the same way, as it
corresponds to a classical Libert scale used in experimental psychology.
But even if the model was constructed using observation spanning in the
interval [1,7], particular audio examples, not considered in the training, can
extend beyond that range. So the interval [0,8] looks like a more probable
range of value for these dimensions and concepts.
The model presented in the ISMIR 2009 paper corresponds more or less to the one
available in version 1.3 of the toolbox. Version 1.3.1 contains major
improvements that would have a significant effect on the emotion model, however
miremotion has not been re-calibrated properly for this new version. I hope we
will offer an update of the emotion model soon.
Best regards,
Olivier
Le 11.1.2011 à 17.30, Mark Mann a écrit :
> Hi
>
> I have a question about miremotion.
>
> The output for a music file in which the ‘Frame’ function has been used
> results in a graph showing the temporal variation of the coefficients for
> happy, sad, etc.
>
> I’ve read the paper and the MIRtoolbox guide, but I still can’t work out what
> this coefficients mean. Does a coefficient of say 6 in sad and 1 in happy
> mean that it is 6 times as happy as sad, or simply much more sad than happy,
> or vice versa?
>
> Thanks and best wishes
>
> Mark Mann
>
> BBC Research & Development
>
>
>
>
> http://www.bbc.co.uk
> This e-mail (and any attachments) is confidential and may contain personal
> views which are not the views of the BBC unless specifically stated.
> If you have received it in error, please delete it from your system.
> Do not use, copy or disclose the information in any way nor act in reliance
> on it and notify the sender immediately.
> Please note that the BBC monitors e-mails sent or received.
> Further communication will signify your consent to this.
Other related posts:
