Grounding language queries in videos aims at identifying the time interval (or moment) semantically relevant to a language query. The solution to this challenging task demands understanding videos' and queries' semantic content and the fine-grained reasoning about their multi-modal interactions. Our key idea is to recast this challenge into an algorithmic graph matching problem. Fueled by recent advances in Graph Neural Networks, we propose to leverage Graph Convolutional Networks to model video and textual information as well as their semantic alignment. To enable the mutual exchange of information across the modalities, we design a novel Video-Language Graph Matching Network (VLG-Net) to match video and query graphs. Core ingredients include representation graphs built atop video snippets and query tokens separately and used to model intra-modality relationships. A Graph Matching layer is adopted for cross-modal context modeling and multi-modal fusion. Finally, moment candidates are created using masked moment attention pooling by fusing the moment's enriched snippet features. We demonstrate superior performance over state-of-the-art grounding methods on three widely used datasets for temporal localization of moments in videos with language queries: ActivityNet-Captions, TACoS, and DiDeMo.
|Original language||English (US)|
|Title of host publication||Proceedings - 2021 IEEE/CVF International Conference on Computer Vision Workshops, ICCVW 2021|
|Publisher||Institute of Electrical and Electronics Engineers Inc.|
|Number of pages||11|
|State||Published - 2021|
|Event||18th IEEE/CVF International Conference on Computer Vision Workshops, ICCVW 2021 - Virtual, Online, Canada|
Duration: Oct 11 2021 → Oct 17 2021
|Name||Proceedings of the IEEE International Conference on Computer Vision|
|Conference||18th IEEE/CVF International Conference on Computer Vision Workshops, ICCVW 2021|
|Period||10/11/21 → 10/17/21|
Bibliographical noteFunding Information:
This paper addresses the problem of text-to-video temporal grounding, where we cast the problem as an algorithmic graph matching. We propose Video-Language Graph Matching Network (VLG-Net) to match the video and language modalities. We represent each modality as graphs and explore four types of edges, Syntactic Edge, Ordering Edge, Semantic Edge, and Matching Edge, to encode local, non-local, and cross-modality relationships to align the video-query pair. Extensive experiments show that our VLG-Net can model inter-and intra-modality context, learn multi-modal fusion and surpass the current state-of-the-art performance on three widely used datasets. Acknowledgments. This work was supported by the King Abdullah University of Science and Technology (KAUST) Office of Sponsored Research through the Visual Computing Center (VCC) funding.
© 2021 IEEE.
ASJC Scopus subject areas
- Computer Vision and Pattern Recognition