Deep convolutional feature transfer across mobile activity recognition domains, sensor modalities and locations
Deep convolutional feature transfer across mobile activity recognition domains, sensor modalities and locations
复制标题
DOI：
10.1145/2971763.2971764
复制
发表时间：
2016-09
期刊：
影响因子：
--
通讯作者：
Francisco Javier Ordonez;D. Roggen
中科院分区：
文献类型：
--
作者：
Francisco Javier Ordonez;D. Roggen
深度卷积网络的卷积层中的内核被认为是特征提取器，逐渐突出上层网络层中更多的特定于域的特征。因此，较低级别的特征可能适合于转移。我们通过在另一个目标域上重用在源域上学习的内核来分析可穿戴活动识别。我们考虑用户之间的传输，应用领域，传感器模态和传感器位置。我们描述了传输各种卷积层沿着模型大小，学习速度，识别性能和训练数据的权衡。通过新颖的内核可视化和比较评估，我们确定了哪些内核主要对传感器特性、运动动力学和身体放置敏感。核转移将训练时间减少了约17%，而不会增加复杂性。我们得出关于何时转移最合适的建议。
Kernels in the convolutional layers of deep convolutional networks are believed to act as feature extractors, progressively highlighting more domain-specific features in the upper network layers. Thus lower-level features might be suitable for transfer. We analyse this in wearable activity recognition by reusing kernels learned on a source domain on another target domain. We consider transfer between users, application domains, sensor modalities and sensor locations. We characterize the trade-offs of transferring various convolutional layers along model size, learning speed, recognition performance and training data. Through novel kernel visualisations and comparative evaluations we identify what kernels are predominantly sensitive to, amongst sensor characteristics, motion dynamics and on-body placement. Kernel transfer reduces training time by ~17% without additional complexity. We derive recommendations on when transfer is most suitable.