Local-Global Video-Text Interactions for Temporal Grounding

TLDR

The study tackles text‑to‑video temporal grounding, aiming to locate the video interval relevant to a text query. We propose a regression‑based model that extracts mid‑level features for semantic phrases, learns bi‑modal interactions between linguistic and visual features across multiple levels, and predicts the target interval by exploiting contextual information from local to global. Ablation studies show that incorporating both local and global context is crucial, and experiments demonstrate that the method surpasses state‑of‑the‑art on Charades‑STA and ActivityNet Captions by 7.44% and 4.61% at Recall@tIoU=0.5.

Abstract

This paper addresses the problem of text-to-video temporal grounding, which aims to identify the time interval in a video semantically relevant to a text query. We tackle this problem using a novel regression-based model that learns to extract a collection of mid-level features for semantic phrases in a text query, which corresponds to important semantic entities described in the query (e.g., actors, objects, and actions), and reflect bi-modal interactions between the linguistic features of the query and the visual features of the video in multiple levels. The proposed method effectively predicts the target time interval by exploiting contextual information from local to global during bi-modal interactions. Through in-depth ablation studies, we find out that incorporating both local and global context in video and text interactions is crucial to the accurate grounding. Our experiment shows that the proposed method outperforms the state of the arts on Charades-STA and ActivityNet Captions datasets by large margins, 7.44% and 4.61% points at Recall@tIoU=0.5 metric, respectively.