Vidoumo reads the captions YouTube already has and adds the reading over every kanji, the English underneath, and a dictionary behind every word you click. A JLPT level in front of the title tells you how hard the video is before you press play.