• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>Sensing and imaging >A Lightweight Two-stream Fusion Deep Neural Network Based on ResNet Model for Sports Motion Image Recognition
【24h】

A Lightweight Two-stream Fusion Deep Neural Network Based on ResNet Model for Sports Motion Image Recognition

机译：基于Resnet模型的轻量级二流融合深神经网络体育型图像识别

获取原文
获取原文并翻译 | 示例
           
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Sports motion recognition aims to track the movement of some key points in thetime domain and record human movement, which is of great significance for competitivetraining and national fitness. However, the traditional motion recognitionmethods have the disadvantages such as low recognition rate, time-consumingdue to the vast model parameters and edge artifacts, etc. By combining the shallowmulti-scale network with the deep network, this paper proposes a novel sportsmotion image recognition based on a lightweight two-stream fusion deep neural networkwith the ResNet model. This model can greatly reduce the number of modelparameters and improve the accuracy of feature extraction. Finally, we conduct theexperiments on four public data sets: UCF101, HMDB51, MSR3D and UCF-Sport.Experiment results show that the proposed algorithm has higher recognition accuracyand stability with various complex actions compared with other state-of-the-artmotion recognition methods.
机译：运动运动识别旨在追踪一些关键点的运动 时域和记录人类运动，这对于竞争具有重要意义 培训和国家健身。 但是，传统运动识别 方法具有低识别率，耗时等缺点 由于组合浅层的巨大模型参数和边缘伪影等 多尺度网络与深网络，本文提出了一种新颖运动 基于轻量级二流融合深神经网络的运动图像识别 使用reset模型。 该模型可以大大减少模型的数量 参数和提高特征提取的准确性。 最后，我们进行 四个公共数据集的实验：UCF101，HMDB51，MSR3D和UCF-Sport。 实验结果表明，该算法具有较高的识别准确性 与各种复杂的动作的稳定性与其他最先进的稳定性相比 运动识别方法。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号