Graph Interaction Networks for Relation Transfer in Human Activity Videos
Graph Interaction Networks for Relation Transfer in Human Activity Videos
复制标题
用于人类活动视频中关系转移的图交互网络
DOI：
10.1109/tcsvt.2020.2973301
复制
发表时间：
2020-02
影响因子：
8.4
通讯作者：
Zhou Jie
中科院分区：
文献类型：
--
作者：
Tang Yansong;Wei Yi;Yu Xumin;Lu Jiwen;Zhou Jie
近年来，在使用图卷积网络（GCN）进行各种基于图的数据丰富的视频分析任务方面取得了快速进展。然而，探索不同图之间的可传递知识，这是一个具有广泛和潜在应用的方向，却很少有人研究。为了解决这个问题，我们提出了一个图交互网络（GINs）模型，用于在两个图之间传输关系知识。与传统的领域自适应或知识蒸馏方法不同，我们的GIN专注于“自学”权重矩阵，这是输入数据的更高级别的表示。权重矩阵的每个元素表示图中不同节点之间的成对关系。此外，我们通过设计特定于任务的损失函数来引导网络在权重矩阵之间传递知识，从而在传递过程中很好地保留了关系信息。我们对两种不同的视频分析场景进行了实验，包括一种新的针对不同数据集的无监督基于动作识别的设置，以及具有多模态输入的监督组活动识别。在六个广泛使用的数据集上进行的大量实验表明，与最先进的GIN相比，我们的GIN具有非常有竞争力的性能。
Recent years have witnessed rapid progress in employing graph convolutional networks (GCNs) for various video analysis tasks where graph-based data abound. However, exploring the transferable knowledge between different graphs, which is a direction with wide and potential applications, has been rarely studied. To address this issue, we propose a graph interaction networks (GINs) model for transferring relation knowledge across two graphs. Different from conventional domain adaptation or knowledge distillation approaches, our GINs focus on a “self-learned” weight matrix, which is a higher-level representation of the input data. And each element of the weight matrix represents the pair-wise relation among different nodes within the graph. Moreover, we guide the networks to transfer the knowledge across the weight matrices by designing a task-specific loss function, so that the relation information is well preserved during transfer. We conduct experiments on two different scenarios for video analysis, including a new proposed setting for unsupervised skeleton-based action recognition across different datasets, and supervised group activity recognition with multi-modal inputs. Extensive experiments on six widely used datasets illustrate that our GINs achieve very competitive performance in comparison with the state-of-the-arts.
登录
查看更多内容
用于场景识别和领域适应的混合 CNN 和基于字典的模型
DOI：
10.1109/tcsvt.2015.2511543
发表时间：
2016-01
影响因子：
8.4
作者：
Guo-Sen Xie;Xu-Yao Zhang;Shuicheng Yan;Cheng-Lin Liu
通讯作者：
Cheng-Lin Liu
DOI：
10.1109/iccvw.2009.5457461
发表时间：
2009-09
期刊：
2009 IEEE 12th International Conference on Computer Vision Workshops, ICCV Workshops
影响因子：
--
作者：
Wongun Choi;Khuram Shahid;S. Savarese
通讯作者：
Wongun Choi;Khuram Shahid;S. Savarese
学习语义——为群体活动识别保留注意力和情境交互
影响因子：
10.6
作者：
Tang Yansong;Lu Jiwen;Wang Zian;Yang Ming;Zhou Jie
通讯作者：
Zhou Jie
影响因子：
3.9
作者：
Mingsheng Long;Zhangjie Cao;Jianmin Wang;Michael I. Jordan
通讯作者：
Mingsheng Long;Zhangjie Cao;Jianmin Wang;Michael I. Jordan
DOI：
10.1007/978-3-030-01264-9_11
发表时间：
2017-11
期刊：
--
影响因子：
--
作者：
Zelun Luo;Jun-Ting Hsieh;Lu Jiang;Juan Carlos Niebles;Li Fei-Fei-Li-Fei-Fei-48004138
通讯作者：
Zelun Luo;Jun-Ting Hsieh;Lu Jiang;Juan Carlos Niebles;Li Fei-Fei-Li-Fei-Fei-48004138