Introduction - If you have any usage issues, please Google them yourself
The samples are classified into positive samples, negative samples, and the samples are difficult to distinguish. The samples are labeled as 1,- 1, and the sample features are calculated by hog, and written into the vsm model in xml format. Test samples, only need to load xml file to calculate the hog feature of the test sample, through the prediction function can be obtained which he belongs to the category.