DeepAVO：基于特征提取的深度视觉里程计的高效姿态优化
DeepAVO: Efficient Pose Refining with Feature Distilling for Deep Visual Odometry
摘要：
随着人们对自动驾驶的兴趣日益浓厚，视觉里程计（VO）通过分析车载摄像头捕获的图像序列来估计移动物体的位置和方向的技术已经得到了很好的研究。本文从深度学习（DL）的角度研究单眼VO。与目前大多数基于学习的方法不同，我们的方法称为DeepAVO，它建立在直觉上，即特征对不同的运动模式有区别的贡献。具体来说，我们提出了一种新的四分支网络，通过利用卷积神经网络（CNN）专注于光流输入的不同象限来学习旋转和平移。为了提高特征选择的能力，我们进一步引入了一个有效的通道空间注意力机制，以迫使每个分支明确提取相关信息的特定帧到帧（F2F）的运动估计。在涉及室外驾驶和室内步行场景的各种数据集上的实验表明，所提出的DeepAVO大大优于最先进的单目方法，表现出与立体VO算法竞争的性能，并验证了推广的潜力。
英文摘要：
The technology for Visual Odometry (VO) that estimates the position and orientation of the moving object through analyzing the image sequences captured by on-board cameras, has been well investigated with the rising interest in autonomous driving. This paper studies monocular VO from the perspective of Deep Learning (DL). Unlike most current learning-based methods, our approach, called DeepAVO, is established on the intuition that features contribute discriminately to different motion patterns. Specifically, we present a novel four-branch network to learn the rotation and translation by leveraging Convolutional Neural Networks (CNNs) to focus on different quadrants of optical flow input. To enhance the ability of feature selection, we further introduce an effective channel-spatial attention mechanism to force each branch to explicitly distill related information for specific Frame to Frame (F2F) motion estimation. Experiments on various datasets involving outdoor driving and indoor walking scenarios show that the proposed DeepAVO outperforms the state-of-the-art monocular methods by a large margin, demonstrating competitive performance to the stereo VO algorithm and verifying promising potential for generalization.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
