MAC: Mining Activity Concepts for Language-Based Temporal Localization

We address the problem of language-based temporal localization in untrimmed videos. Compared to temporal localization with fixed categories, this problem is more challenging as the language-based queries not only have no pre-defined activity list but also may contain complex descriptions. Previous methods address the problem by considering features from video sliding windows and language queries and learning a subspace to encode their correlation, which ignore rich semantic cues about activities in videos and queries. We propose to mine activity concepts from both video and language modalities by applying the actionness score enhanced Activity Concepts based Localizer (ACL). Specifically, the novel ACL encodes the semantic concepts from verb-obj pairs in language queries and leverages activity classifiers' prediction scores to encode visual concepts. Besides, ACL also has the capability to regress sliding windows as localization results. Experiments show that ACL significantly outperforms state-of-the-arts under the widely used metric, with more than 5% increase on both Charades-STA and TACoS datasets.

1. Introduction

We address natural language-based temporal localization in untrimmed videos. Given an untrimmed video and a natural language query, natural language-based temporal localization aims to determine the start and end times of the activities in the query. Compared to the activity temporal localization task [25], [12], [33], [34], [3], the natural language-based temporal localization has two unique characteristics. First, the natural language-based temporal localization is not constrained by a pre-defined activity label list; second, the language query can be more complex and may contain multiple activities. e.g. “person begin opening the refrigerator to find more food”.