论文标题
无监督的流程细化近运动边界
Unsupervised Flow Refinement near Motion Boundaries
论文作者
论文摘要
基于深度学习的无监督光流估计器由于对地面真理的成本和难度而引起了越来越多的关注。尽管多年来通过平均终点误差（EPE）衡量的性能有所改善，但沿运动边界（MBS）的流量估计仍然较差，在运动边界（MBS）中，流动不平滑，通常假定的流动不平滑，而神经网络计算得出的功能则被多个运动污染。为了改善无监督的设置中的流量，我们设计了一个框架，该框架通过分析沿边界候选者的视觉变化并用更远的动作来代替接近检测的动作来检测MBS。我们提出的算法比具有相同输入的基线方法更准确地检测边界，并且可以改善任何流动预测变量的估计值，而无需额外的训练。
Unsupervised optical flow estimators based on deep learning have attracted increasing attention due to the cost and difficulty of annotating for ground truth. Although performance measured by average End-Point Error (EPE) has improved over the years, flow estimates are still poorer along motion boundaries (MBs), where the flow is not smooth, as is typically assumed, and where features computed by neural networks are contaminated by multiple motions. To improve flow in the unsupervised settings, we design a framework that detects MBs by analyzing visual changes along boundary candidates and replaces motions close to detections with motions farther away. Our proposed algorithm detects boundaries more accurately than a baseline method with the same inputs and can improve estimates from any flow predictor without additional training.
