Hi Adam,

Go through these articles on different techniques used to improve the accuracy of both video and image segementation.

https://arxiv.org/pdf/1605.00732

www.mirlab.org/conference_papers/International_Conference/.../0001165.pdf

juew.org/publication/mattingSurvey.pdf

Best!

Thank you for that.

Best regards

Will try Adam from algorithm perspective : There are developments in the field of :

https://cvfxbook.com/useful-code/

It probably works with more than 2 techniques eg- Alpha /graph based / other ML algos

http://vixra.org/pdf/1308.0085v1.pdf

Thank you @Aparna for these useful URL.

Best Regards

Funny thing Adam , most of the algorithms are "multi-modal" per se ...

e.g:  SV clustering and graph based links .

Thank you @Aparna Sathya Murthy. I appreciate your help.

Indeed, May you reas my artcile is spectral matting and give me what you think about ? 

Best Regards,

Article Gradient Color Tensor based Approach for Spectral Matting

Will do , thank you for sharing Adam

Not at all.

I'am waiting you'r remarks. 

Remarks Adam , it is well thought thru work ...

I specifically like the "small windowing" idea.  The tensor space analysis (GAMP metrics) for color discontinuity.

The color model is for one specific color space ? I mean as you have mentioned RGB  !

How have you based, delta  and gamma (just inquisitive )

I have used Fuzzy-C means , there also we claim "empirical"  , here it is # based unlike your case which should be color metric ??