VideoMCC: a New Benchmark for Video Comprehension

TL;DR

VideoMCC introduces a new benchmark for evaluating high-level semantic understanding in videos through a well-defined captioning task with an interpretable performance measure, addressing limitations of previous methods.

Contribution

The paper formulates VideoMCC as a new task, provides a semi-automatic benchmark creation method, releases a large-scale dataset, and evaluates approaches to understand challenges in video comprehension.

Findings

Established a new video captioning benchmark with clear metrics

Created a large-scale, publicly available video dataset

Assessed various approaches to identify key challenges in video understanding

Abstract

While there is overall agreement that future technology for organizing, browsing and searching videos hinges on the development of methods for high-level semantic understanding of video, so far no consensus has been reached on the best way to train and assess models for this task. Casting video understanding as a form of action or event categorization is problematic as it is not fully clear what the semantic classes or abstractions in this domain should be. Language has been exploited to sidestep the problem of defining video categories, by formulating video understanding as the task of captioning or description. However, language is highly complex, redundant and sometimes ambiguous. Many different captions may express the same semantic concept. To account for this ambiguity, quantitative evaluation of video description requires sophisticated metrics, whose performance scores are typically hard to interpret by humans. This paper provides four contributions to this problem. First, we formulate Video Multiple Choice Caption (VideoMCC) as a new well-defined task with an easy-to-interpret performance measure. Second, we describe a general semi-automatic procedure to create benchmarks for this task. Third, we publicly release a large-scale video benchmark created with an implementation of this procedure and we include a human study that assesses human performance on our dataset. Finally, we propose and test a varied collection of approaches on this benchmark for the purpose of gaining a better understanding of the new challenges posed by video comprehension.