This forum is disabled, please visit https://forum.opencv.org
| 2015-08-13 03:09:43 -0600 | answered a question | Ideas for separating LDA clusters to improve SVM accuracy? I suggest you use feature extractor first instead of raw pixels. feature with LDA are much better than just raw pixels. Have a look of this paper: file:///home/tony/Downloads/nicta_publication_slides_8868.pdf |