TGIF-QA: Toward Spatio-Temporal Reasoning in Visual Question Answering
This work addresses the gap in VQA research by moving from images to videos, enabling spatio-temporal reasoning for applications in video understanding, though it is incremental as it builds on existing VQA techniques.
The paper tackles the problem of extending visual question answering (VQA) to videos by introducing new tasks requiring spatio-temporal reasoning, and it presents a new dataset and a dual-LSTM approach with spatial and temporal attention, showing effectiveness over conventional methods through empirical evaluations.
Vision and language understanding has emerged as a subject undergoing intense study in Artificial Intelligence. Among many tasks in this line of research, visual question answering (VQA) has been one of the most successful ones, where the goal is to learn a model that understands visual content at region-level details and finds their associations with pairs of questions and answers in the natural language form. Despite the rapid progress in the past few years, most existing work in VQA have focused primarily on images. In this paper, we focus on extending VQA to the video domain and contribute to the literature in three important ways. First, we propose three new tasks designed specifically for video VQA, which require spatio-temporal reasoning from videos to answer questions correctly. Next, we introduce a new large-scale dataset for video VQA named TGIF-QA that extends existing VQA work with our new tasks. Finally, we propose a dual-LSTM based approach with both spatial and temporal attention, and show its effectiveness over conventional VQA techniques through empirical evaluations.