TutorialVQA: Question Answering Dataset for Tutorial Videos
This addresses a gap in video QA for instructional content, but it is incremental as it focuses on a specific domain (image editing tutorials).
The authors tackled the lack of datasets for multi-step and non-factoid question answering in videos by introducing TutorialVQA, a dataset of about 6,000 triples from screencast tutorials, and found that baseline algorithms using transcripts show the task is challenging.
Despite the number of currently available datasets on video question answering, there still remains a need for a dataset involving multi-step and non-factoid answers. Moreover, relying on video transcripts remains an under-explored topic. To adequately address this, We propose a new question answering task on instructional videos, because of their verbose and narrative nature. While previous studies on video question answering have focused on generating a short text as an answer, given a question and video clip, our task aims to identify a span of a video segment as an answer which contains instructional details with various granularities. This work focuses on screencast tutorial videos pertaining to an image editing program. We introduce a dataset, TutorialVQA, consisting of about 6,000manually collected triples of (video, question, answer span). We also provide experimental results with several baselines algorithms using the video transcripts. The results indicate that the task is challenging and call for the investigation of new algorithms.