TGIF-QA: Toward Spatio-Temporal Reasoning in Visual Question Answering

TL;DR

This paper introduces TGIF-QA, a new dataset and tasks for video visual question answering that require spatio-temporal reasoning, along with a dual-LSTM model with attention mechanisms demonstrating improved performance.

Contribution

It extends VQA to videos with new tasks, provides a large-scale dataset, and proposes a dual-LSTM model with spatial and temporal attention for better reasoning.

Findings

The dual-LSTM model outperforms traditional VQA methods on video tasks.

TGIF-QA dataset enables research on spatio-temporal reasoning in videos.

New tasks challenge models to understand both spatial and temporal aspects of videos.

Abstract

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.