Localizing Visual Sounds the Hard Way
Honglie Chen, Weidi Xie, Triantafyllos Afouras, Arsha Nagrani, Andrea Vedaldi, Andrew Zisserman
2021年份
90顶会引用
摘要
We localise sound sources in videos without manual annotation. Our key contribution is an automatic negative mining technique through differentiable thresholding of a cross-modal correspondence score map into a Tri-map. We use background regions with low correlation to the given sound as 'hard negatives' in a contrastive learning framework.
问问这篇 Paper
智能体会读完全文。
Lune 把这篇 Paper 索引到了每一个公式，引用它的顶会 Paper 也一样。你提问，回答直接引用原文。
引用它的顶会 Paper90
- Attention Bottlenecks for Multimodal FusionArsha Nagrani, Shan Yang, Anurag Arnab, Aren Jansen 等NeurIPS 2021 · 被引用 884 次
- Learning to Answer Questions in Dynamic Audio-Visual ScenariosGuangyao Li, Yake Wei, Yapeng Tian, Chenliang Xu 等CVPR 2022 · 被引用 101 次
- AVSegFormer: Audio-Visual Segmentation with TransformerShengyi Gao, Zhe Chen, Guo Chen, Wenhai Wang 等AAAI 2024 · 被引用 96 次
- A Closer Look at Weakly-Supervised Audio-Visual Source LocalizationShentong Mo, Pedro MorgadoNeurIPS 2022 · 被引用 92 次
- Multi-modal Grouping Network for Weakly-Supervised Audio-Visual Video ParsingShentong Mo, Yapeng TianNeurIPS 2022 · 被引用 73 次
它引用的顶会 Paper2
相关 Paper
- Learning Audio-Visual Source Localization via False Negative Aware Contrastive LearningWeixuan Sun, Jiayi Zhang, Jianyuan Wang, Zheyuan Liu 等CVPR 2023
- Cross-Modal Label Contrastive Learning for Unsupervised Audio-Visual Event LocalizationPeijun Bao, Wenhan Yang, Boon Poh Ng, Meng Hwa Er 等AAAI 2023 · 被引用 13 次
- Unsupervised Sounding Pixel LearningYining Zhang, Yanli Ji, Yang YangEMNLP 2023 · 被引用 2 次
- A Proposal-based Paradigm for Self-supervised Sound Source Localization in VideosHanyu Xuan, Zhiliang Wu, Jian Yang, Yan Yan 等CVPR 2022 · 被引用 27 次
- Weakly-Supervised Audio-Visual SegmentationShentong Mo, Bhiksha RajNeurIPS 2023 · 被引用 26 次
