Semi-Supervised Learning of Optical Flow by Flow Supervisor
This addresses the challenge of reducing annotation effort for optical flow estimation in computer vision, though it is incremental as it builds on existing self-supervision methods.
The paper tackles the problem of fine-tuning optical flow CNNs without ground-truth flows by proposing a flow supervisor method for self-supervision, achieving meaningful improvements over state-of-the-art models on Sintel and KITTI benchmarks.
A training pipeline for optical flow CNNs consists of a pretraining stage on a synthetic dataset followed by a fine tuning stage on a target dataset. However, obtaining ground truth flows from a target video requires a tremendous effort. This paper proposes a practical fine tuning method to adapt a pretrained model to a target dataset without ground truth flows, which has not been explored extensively. Specifically, we propose a flow supervisor for self-supervision, which consists of parameter separation and a student output connection. This design is aimed at stable convergence and better accuracy over conventional self-supervision methods which are unstable on the fine tuning task. Experimental results show the effectiveness of our method compared to different self-supervision methods for semi-supervised learning. In addition, we achieve meaningful improvements over state-of-the-art optical flow models on Sintel and KITTI benchmarks by exploiting additional unlabeled datasets. Code is available at https://github.com/iwbn/flow-supervisor.