difference between "trainImag​eCategoryC​lassifier" and "classificationLearner"

Hi,

If i have some input images/data and i am using the "BagOfFeatures"-Function. Now I can either evaluate with the CassificationLearner and choose a method which fits best and create a Model for further processing within the tool itself or use the "trainImageCategoryClassifier"-Function.

Question is: What is the difference between those two methods? Does the first just uses SVM-method? ClassificationLearner can handle data formated as tables. How about the "trainImageCategoryClassifier"-Function?

Regards

Sven