1

I'm new to machine learning .I'm trying to find a way to detect the existence of handwritten signature in scanned documents. After doing a lot of research ,I found out ,that is essential to use SVM classifiers. So,when there's machine learning ,there have to be training phase and the need for training datasets.I happened to get my hand on a dataset from a signature competition . So now , Do I use segmantation on my image and then use svm Or more simple than : I have 25 documents and I label them as signed not signed ?

Meta Man
  • 21
  • 2
  • We're still learning about ML. Much of the field is experimentation. The basic answer is to try both and see what results you get. If your personal research hasn't already turned up a definitive answer, chances are that there is no such answer for your specific application. – Prune Jul 11 '17 at 18:06

0 Answers0