Efficient estimation for shared latent space using multi-layer perceptron
Efficient estimation for shared latent space using multi-layer perceptron
复制标题
DOI：
10.1109/smc.2017.8122986
复制
发表时间：
2017-10
期刊：
影响因子：
--
通讯作者：
Mariho Ohyama;I. Kobayashi
中科院分区：
文献类型：
--
作者：
Mariho Ohyama;I. Kobayashi
由于时间序列中有很多高维的数据是同时出现的，如嘴唇运动、语音、面部表情等，在捕捉不同维数的时间序列数据之间的对应关系时，需要将这些数据的维数统一起来，以便进行比较。为了实现这一点，高斯过程潜变量模型（GPLVM）通常用于减少高维时间序列数据的大小。在这项研究中，我们提出了一种方法来引入MLP的GPLVM为基础的方法估计潜在的状态。我们将所提出的方法应用于GPLVM，SharedGPLVM，GPDM和SharedGPDM，然后证实我们的方法在效率和精确估计方面优于传统方法。
There are quite a few high dimensional time-series data co-ocurring each other such as lip motions, voices, and face appearances and so on. When capturing the correspondent relationships among those time-series data with different dimensionality, we need to make the dimensionality all the same size so that they can be compared each other. To achieve this, Gaussian Process Latent Variable Models (GPLVM) is often used to reduce the size of high dimensional time-series data. In this study, we propose a method to introduce MLP to GPLVM-based methods in estimating latent states. We applied the proposed method to GPLVM, SharedGPLVM, GPDM, and SharedGPDM, and then confirmed that our method outperforms the conventional methods in terms of efficiency and precisely estimation.