When training pictures you write the translation into a text field or select a picture. There is also a multiple choice mode where you can pick one of three answers. You can configure whether you want to be examined from picture to word or from word to picture or in both directions. You can input dictionaries yourself. There are four demo dictionaries in addition.