Multi-Task Learning with Group Information for Human Action Recognition

机译：具有小组信息的多任务学习，用于人类动作识别

获取原文

页面导航

摘要
著录项
相似文献
相关主题

摘要

Human action recognition is an important and challenging task in computer vision research, due to the variations in human motion performance, interpersonal differences and recording settings. In this paper, we propose a novel multi-task learning framework with group information (MTL-GI) for accurate and efficient human action recognition. Specifically, we firstly obtain group information through calculating the mutual information according to the latent relationship between Gaussian components and action categories, and clustering similar action categories into the same group by affinity propagation clustering. Additionally, in order to explore the relationships of related tasks, we incorporate group information into multi-task learning. Experimental results evaluated on two popular benchmarks (UCF50 and HMDB51 datasets) demonstrate the superiority of our proposed MTL-GI framework.

机译：由于人类运动表现，人际差异和记录设置的变化，人类动作识别在计算机视觉研究中是一项重要且具有挑战性的任务。在本文中，我们提出了一种具有组信息（MTL-GI）的新颖的多任务学习框架，用于准确有效的人类动作识别。具体而言，首先，根据高斯分量与动作类别之间的潜在关系，通过计算互信息，并通过亲和力传播聚类，将相似的动作类别聚类为同一组，从而获得群组信息。此外，为了探索相关任务的关系，我们将小组信息纳入了多任务学习中。在两个流行的基准（UCF50和HMDB51数据集）上评估的实验结果证明了我们提出的MTL-GI框架的优越性。

著录项

来源
《International conference on graphic and image processing》|2017年|106150W.1-106150W.9|共9页
会议地点
作者
Li Qian; Song Wu; Nan Pu; Shulin Xu; Guoqiang Xiao;
展开▼
作者单位

展开▼
会议组织
原文格式 PDF
正文语种
中图分类
关键词
multi-task learning; Fisher vector; mutual information; improved dense trajectories;

机译：多任务学习;费舍尔向量;相互信息;改进的密集轨迹;

相似文献

外文文献
中文文献
专利

1. Multi-Domain and Multi-Task Learning for Human Action Recognition [J] . An-An Liu, Ning Xu, Wei-Zhi Nie, IEEE Transactions on Image Processing . 2019,第2期

机译：多域多任务学习的人类动作识别
2. Hierarchical Clustering Multi-Task Learning for Joint Human Action Grouping and Recognition [J] . An-An Liu, Yu-Ting Su, Wei-Zhi Nie, IEEE Transactions on Pattern Analysis and Machine Intelligence . 2017,第1期

机译：分层聚类多任务学习，用于联合人类行为分组和识别
3. Human action recognition via multi-task learning base on spatial-temporal feature [J] . Guo Wenzhong, Chen Guolong Information Sciences: An International Journal . 2015,第Null期

机译：基于时空特征的多任务学习对人类动作的识别
4. Multi-Task Learning with Group Information for Human Action Recognition [C] . Li Qian, Song Wu, Nan Pu, International Conference on Graphic and Image Processing . 2017

机译：具有人类行动认可的组信息的多任务学习
5. Multi-task learning deep neural networks for automatic speech recognition [D] . Chen, Dongpeng. 2015

机译：多任务学习深度神经网络自动语音识别
6. Semi-supervised multi-task learning for predicting interactions between HIV-1 and human proteins [O] . Yanjun Qi, Oznur Tastan, Jaime G. Carbonell, -1

机译：半监督多任务学习用于预测HIV-1和人类蛋白质之间的相互作用
7. Multi-task Deep Learning for Real-Time 3D Human Pose Estimation and Action Recognition [O] . Diogo Luvizon, David Picard, Hedi Tabia 2020

机译：实时3D人类姿态估算与动作识别的多任务深度学习
8. Evaluation of an Integrated Multi-Task Machine Learning System with Humans in the Loop [R] . Steinfield, A., Bennett, S. R., Cunningham, K., 2007

机译：人与人之间的综合多任务机器学习系统评估

Multi-Task Learning with Group Information for Human Action Recognition

摘要

著录项

相似文献

相关主题

期刊订阅