Weakly-Supervised Multi-Level Attentional Reconstruction Network for Grounding Textual Queries in Videos
arXiv:2003.07048
Abstract
The task of temporally grounding textual queries in videos is to localize one video segment that semantically corresponds to the given query. Most of the existing approaches rely on segment-sentence pairs (temporal annotations) for training, which are usually unavailable in real-world scenarios. In this work we present an effective weakly-supervised model, named as Multi-Level Attentional Reconstruction Network (MARN), which only relies on video-sentence pairs during the training stage. The proposed method leverages the idea of attentional reconstruction and directly scores the candidate segments with the learnt proposal-level attentions. Moreover, another branch learning clip-level attention is exploited to refine the proposals at both the training and testing stage. We develop a novel proposal sampling mechanism to leverage intra-proposal information for learning better proposal representation and adopt 2D convolution to exploit inter-proposal clues for learning reliable attention map. Experiments on Charades-STA and ActivityNet-Captions datasets demonstrate the superiority of our MARN over the existing weakly-supervised methods.
References in corpus (3)
Cited by in corpus (8)
- Temporal Sentence Grounding in Videos: A Survey and Future Directions
- A Closer Look at Temporal Sentence Grounding in Videos: Dataset and Metric
- Video Moment Retrieval from Text Queries via Single Frame Annotation
- SCANet: Scene Complexity Aware Network for Weakly-Supervised Video Moment Retrieval
- Gaussian Mixture Proposals with Pull-Push Learning Scheme to Capture Diverse Events for Weakly Supervised Temporal Video Grounding
- Variational Cross-Graph Reasoning and Adaptive Structured Semantics Learning for Compositional Temporal Grounding
- Towards Debiasing Temporal Sentence Grounding in Video
- Self-supervised Learning for Semi-supervised Temporal Language Grounding