LearningfromInside: Self-drivenSiameseSampling andReasoningforVideoQuestionAnswering

By inferring the correct answers for video-based questions, video question answering (VideoQA) has attracted increasing research attention due to its huge application potential, as a fundamental technique for vision-to-language reasoning. The task involves acquisition and manipulation of spatio-temporal visual representations guided by the compositional semantics of the linguistic clues[32,15,21,34]. Existingworkscanroughly be divided into two aspects.