WPO-Net: Windowed Pose Optimization Network for Monocular Visual Odometry Estimation.

WPO-Net: Windowed Pose Optimization Network for Monocular Visual Odometry Estimation.
复制标题

DOI:
10.3390/s21238155
复制
发表时间:
2021-12-06
期刊:
Sensors (Basel, Switzerland)
影响因子:
--
通讯作者:
Su S
Su S
中科院分区:
其他
文献类型:
--
作者:
Gadipudi N;Elamvazuthi I;Lu CK;Paramasivam S;Su S

文献摘要

参考文献

被引文献

相似文献

视觉里程计是用于自动驾驶的在3维空间中估计相机的增量定位的过程。已经出现了新的基于学习的方法,这些方法不需要相机校准,并且对外部噪声具有鲁棒性。在这项工作中,提出了一种新的方法,不需要摄像机校准称为“窗口位姿优化网络”来估计单目摄像机的6个自由度的位姿。所提出的网络的架构是基于监督学习的方法与特征编码器和姿态回归器,采取多个连续的两个灰度图像堆栈在每一步的训练和强制执行的复合姿态约束。KITTI数据集被用来评估所提出的方法的性能。该方法产生的旋转误差为3.12度/100米,训练时间为41.32毫秒,而推理时间为7.87毫秒。实验证明,该方法的竞争力的性能,以其他国家的最先进的相关作品,这表明所提出的技术的新奇。
Visual odometry is the process of estimating incremental localization of the camera in 3-dimensional space for autonomous driving. There have been new learning-based methods which do not require camera calibration and are robust to external noise. In this work, a new method that do not require camera calibration called the “windowed pose optimization network” is proposed to estimate the 6 degrees of freedom pose of a monocular camera. The architecture of the proposed network is based on supervised learning-based methods with feature encoder and pose regressor that takes multiple consecutive two grayscale image stacks at each step for training and enforces the composite pose constraints. The KITTI dataset is used to evaluate the performance of the proposed method. The proposed method yielded rotational error of 3.12 deg/100 m, and the training time is 41.32 ms, while inference time is 7.87 ms. Experiments demonstrate the competitive performance of the proposed method to other state-of-the-art related works which shows the novelty of the proposed technique.
DOI: 10.1109/tpami.2017.2658577
发表时间: 2018-03-01
影响因子: 23.6
作者:
Engel, Jakob;Koltun, Vladlen;Cremers, Daniel
通讯作者: Cremers, Daniel
DOI: 10.1109/tits.2021.3053412
发表时间: 2021-02-02
影响因子: 8.5
作者:
Liu, Yiling;Wang, Hesheng;Wang, Xinlei
通讯作者: Wang, Xinlei
DOI: 10.3390/s21144734
发表时间: 2021-07-11
期刊: Sensors (Basel, Switzerland)
影响因子: --
作者:
Mazurek P;Hachaj T
通讯作者: Hachaj T
DOI: 10.1109/tro.2015.2463671
发表时间: 2015-10-01
影响因子: 7.8
作者:
Mur-Artal, Raul;Montiel, J. M. M.;Tardos, Juan D.
通讯作者: Tardos, Juan D.
DOI: 10.1002/rob.20103
发表时间: 2006-01-01
影响因子: 8.3
作者:
Nister, David;Naroditsky, Oleg;Bergen, James
通讯作者: Bergen, James