arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2403.19294cs.CVcs.LG
收藏

FlowDepth:面向自监督单目深度估计的光流解耦方法

FlowDepth: Decoupling Optical Flow for Self-Supervised Monocular Depth Estimation

College of Electronical and Information Engineering, Tongji University(同济大学电子与信息工程学院)

Yiyang Sun, Zhiyuan Xu, Xiaonian Wang, Jing Yao

AI总结:

FlowDepth 通过动态运动光流模块解耦光流以处理运动物体失配,并结合深度线索感知模糊与代价体稀疏损失缓解高频和低纹理区域的光度误差不公平性,在 KITTI 和 Cityscapes 上取得领先的深度估计性能。

摘要:

自监督多帧方法目前在深度估计中已取得有前景的结果。然而,这些方法常因运动物体破坏静态假设而遭受失配问题。此外,在图像的高频或低纹理区域计算光度误差时可能出现不公平性。为解决这些问题,现有方法使用额外的语义先验黑盒网络来分离运动物体,并且仅在损失层面改进模型。因此,我们提出 FlowDepth,其中动态运动光流模块(DMFM)通过基于机制的途径解耦光流,并对动态区域进行 warp,从而解决失配问题。针对高频和低纹理区域引起的光度误差不公平性,我们分别在输入层面和损失层面使用深度线索感知模糊(DCABlur)和代价体稀疏损失来解决问题。在 KITTI 和 Cityscapes 数据集上的实验结果表明,我们的方法优于当前最先进的方法。

英文摘要:

Self-supervised multi-frame methods have currently achieved promising results in depth estimation. However, these methods often suffer from mismatch problems due to the moving objects, which break the static assumption. Additionally, unfairness can occur when calculating photometric errors in high-freq or low-texture regions of the images. To address these issues, existing approaches use additional semantic priori black-box networks to separate moving objects and improve the model only at the loss level. Therefore, we propose FlowDepth, where a Dynamic Motion Flow Module (DMFM) decouples the optical flow by a mechanism-based approach and warps the dynamic regions thus solving the mismatch problem. For the unfairness of photometric errors caused by high-freq and low-texture regions, we use Depth-Cue-Aware Blur (DCABlur) and Cost-Volume sparsity loss respectively at the input and the loss level to solve the problem. Experimental results on the KITTI and Cityscapes datasets show that our method outperforms the state-of-the-art methods.

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。

↑