Self-supervised AutoFlow
This addresses the need for optical flow methods that work on real-world videos without ground truth labels, though it appears incremental as it builds directly on AutoFlow.
The paper tackles the problem of AutoFlow requiring ground truth labels for optical flow training by introducing self-supervised AutoFlow, which uses self-supervised loss as a search metric instead. The result is that it performs on par with AutoFlow on Sintel and KITTI datasets and better on the real-world DAVIS dataset, with competitive results in semi-supervised settings.
Recently, AutoFlow has shown promising results on learning a training set for optical flow, but requires ground truth labels in the target domain to compute its search metric. Observing a strong correlation between the ground truth search metric and self-supervised losses, we introduce self-supervised AutoFlow to handle real-world videos without ground truth labels. Using self-supervised loss as the search metric, our self-supervised AutoFlow performs on par with AutoFlow on Sintel and KITTI where ground truth is available, and performs better on the real-world DAVIS dataset. We further explore using self-supervised AutoFlow in the (semi-)supervised setting and obtain competitive results against the state of the art.