• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>International Conference on Artificial Intelligence and Security >Multiple Obstacle Detection for Assistance Driver System Using Deep Neural Networks
【24h】

Multiple Obstacle Detection for Assistance Driver System Using Deep Neural Networks

机译：深度神经网络的辅助驾驶员系统多障碍物检测

获取原文
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Multiple obstacle detection is a challenging problem and has many important applications including video tracking, intelligence surveillance, robot navigation and autonomous driving. In existing methods, individual obstacle's detection and contextual visual patterns are modeled separately and interactions from obstacles and their surroundings are mostly considered in a symmetric way, which we argue is not an optimal strategy. To tackle these difficulties, in this paper, we propose a deep convolutional networks for solving the online multiple obstacles detection problem. The method consists of deep visual information extraction and visual pattern learning. They are modeled as deep Convolution Neural Networks, which are able to learn discriminative visual features for obstacle detection and model inter-object relations in an asymmetric way and give the orientation extraction for the moving obstacles. The deep learning framework is trained in an end-to-end manner for better adapting the influences of visual information as well as inter-object relations and orientation information. Extensive experimental comparisons with state-of-the-arts as well as detailed component analysis of the proposed method on the benchmarks demonstrate the effectiveness of our proposed framework.
机译：多障碍检测是一个具有挑战性的问题，具有许多重要的应用，包括视频跟踪，智能监视，机器人导航和自动驾驶。在现有方法中，分别对单个障碍物的检测和上下文视觉模式进行建模，并且大多以对称方式考虑与障碍物及其周围环境的交互，因此我们认为这不是最佳策略。为了解决这些困难，本文提出了一种深度卷积网络，用于解决在线多障碍物检测问题。该方法包括深度视觉信息提取和视觉模式学习。它们被建模为深层卷积神经网络，能够学习判别性视觉特征以进行障碍物检测，并以非对称方式对物体之间的关系进行建模，并提供运动障碍物的方向提取。深度学习框架以端到端的方式进行训练，以更好地适应视觉信息以及对象间关系和方向信息的影响。与最新技术进行了广泛的实验比较，并在基准上对提出的方法进行了详细的组件分析，证明了我们提出的框架的有效性。

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