We present an unsupervised learning approach for optical flow estimation by improving the upsampling and learning of pyramid network. We design a self-guided upsample module to tackle the interpolation blur problem caused by bilinear upsampling between pyramid levels. Moreover, we propose a pyramid distillation loss to add supervision for intermediate levels via distilling the finest flow as pseudo labels. By integrating these two components together, our method achieves the best performance for unsupervised optical flow learning on multiple leading benchmarks, including MPI-SIntel, KITTI 2012 and KITTI 2015. In particular, we achieve EPE=1.4 on KITTI 2012 and F1=9.38% on KITTI 2015, which outperform the previous state-of-the-art methods by 22.2% and 15.7%, respectively.

翻译：我们通过改进金字塔网络的增殖和学习,为光学流量估算提供了一个不受监督的学习方法。我们设计了一个自导上流模型,以解决金字塔层之间双线上层取样造成的内插模糊问题。此外,我们提出一个金字塔蒸馏损失,通过将最佳流水蒸馏成假标签,增加中间层的监管。通过将这两个组成部分结合起来,我们的方法在多种主要基准,包括MPI-SIntel、KITTI 2012和KITTI 2015上取得了最佳的无监控光流学习,特别是,我们实现了2012年KITTI的EPE=1.4和2015年KITTI的F1=9.38%的EPE=1.4,这分别比以往最先进的方法高出22.2%和15.7%。