首页> 外国专利> LEARNING METHOD AND LEARNING DEVICE AND TESTING METHOD AND TESTING DEVICE FOR DETECTING PARKING SPACES BY USING POINT REGRESSION RESULTS AND RELATIONSHIP BETWEEN POINTS TO THEREBY PROVIDE AN AUTO-PARKING SYSTEM

LEARNING METHOD AND LEARNING DEVICE AND TESTING METHOD AND TESTING DEVICE FOR DETECTING PARKING SPACES BY USING POINT REGRESSION RESULTS AND RELATIONSHIP BETWEEN POINTS TO THEREBY PROVIDE AN AUTO-PARKING SYSTEM

机译:利用点回归结果以及点与点之间的关系来检测停车位的学习方法,学习装置以及测试方法和测试装置,提供了一种自动配对系统

摘要

The present invention provides a learning method for detecting the parking space using a probability distribution for a determination point of a parking space and a relational linear segment information on a relationship between the determination points (a ) The learning device calculates each of one or more predicted probability distributions for each of the determination points by applying (i) a first CNN (Convolutional Neural Network) to the parking situation image by applying a first CNN regression operation And (ii) performing a process of causing a second CNN to generate predictive relationship linear segment information by applying a second CNN regression operation to the parking situation image; And (b) the learning apparatus causing the loss layer to (i) perform a backpropagation using the first loss to learn the parameters of the first CNN and (ii) the second loss. And performing a process of learning the parameters of the second CNN by performing backpropagation using the method.
机译:本发明提供了一种用于使用停车位的确定点的概率分布以及关于确定点之间的关系的线性关系段信息来检测停车位的学习方法。通过(i)通过应用第一CNN回归操作将第一CNN(卷积神经网络)应用于停车状况图像和(ii)执行使第二CNN生成线性预测关系的过程来确定每个确定点的概率分布通过将第二CNN回归操作应用于停车状况图像来分割信息;并且(b)使损失层(i)使用第一损失执行反向传播以学习第一CNN的参数和(ii)第二损失的学习装置。并且通过使用该方法执行反向传播来执行学习第二CNN的参数的过程。

著录项

相似文献

  • 专利
  • 外文文献
  • 中文文献
获取专利

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