首页>
外国专利>
APPARATUS AND METHOD FOR UNSUPERVISED DOMAIN ADAPTATION
APPARATUS AND METHOD FOR UNSUPERVISED DOMAIN ADAPTATION
展开▼
机译:用于无监督域适应的装置和方法
展开▼
页面导航
摘要
著录项
相似文献
摘要
An apparatus and method for adapting an unsupervised domain are provided. An unsupervised domain adaptation apparatus according to an exemplary embodiment adapts a deep learning model in which supervised learning for a source domain is completed to an unsupervised domain adaptation to a target domain. An unsupervised domain adaptation apparatus, comprising: a plurality of first data belonging to the source domain (x si ) and a pair of labels (y si ) for each of the first data (x si , y si ), A forward pass is performed by inputting a plurality of second data (x Tj ) belonging to the target domain into the deep learning model, respectively, and a Bernoulli distribution, which is a trial probability p, during the execution of the forward pass. A first learning unit for inserting a dropout following) into the deep learning model; And an uncertainty vector for each class-specific predicted value (y-prediction) and the label (y si ) output through the forward pass, and the second data (x Tj) output through the forward pass. And a second learning unit that performs back propagation to minimize uncertainty about a learning parameter of the deep learning model by making each input.
展开▼
机译:提供了一种用于调整无监督域的装置和方法。根据示例性实施例的无监督域自适应装置适应深度学习模型,其中对源域的监督学习被完成到对目标域的无监督域自适应。一个无监督的域适配装置,包括:属于源域的多个第一数据(x s i sup>)和一对标签(ys sub > i sup>)对于第一个数据(xs sub> i sup>,y s sub> i sup >),通过将属于目标域的多个第二数据(x t sub> j sup>)分别在深度学习模型中输入到深度学习模型来执行前向通过,以及Bernoulli在执行前进通过期间,分布是试验概率P.一个用于将辍学的第一学习单元进入深度学习模型;以及通过前向通行证输出的每个类特定的预测值(y-prediction)和标签(y s sub> i i i i i i i i i i i i i i i i i i i i i i i i i i i i i )(x t sub> j)通过前向通过输出。 sup>和第二学习单元,其执行反向传播以通过制造每个输入来最小化关于深度学习模型的学习参数的不确定性。
展开▼