Publications by authors named "Wending Yan"
Adverse Weather Optical Flow: Cumulative Homogeneous-Heterogeneous Adaptation.
IEEE Trans Pattern Anal Mach Intell
September 2024
Optical flow has made great progress in clean scenes, while suffers degradation under adverse weather due to the violation of the brightness constancy and gradient continuity assumptions of optical flow. Typically, existing methods mainly adopt domain adaptation to transfer motion knowledge from clean to degraded domain through one-stage adaptation. However, this direct adaptation is ineffective, since there exists a large gap due to adverse weather and scene style between clean and real degraded domains.
View Article and Find Full Text PDFArticle Synopsis
- * The proposed method introduces a two-stage video-based approach: the first stage uses a single image module to create initial clean results, while the second stage refines these results using multiple frames to ensure temporal consistency.
- * By employing techniques like optical flow and deformable convolution layers for frame alignment and utilizing unsupervised learning for training, the method shows improved performance in removing raindrops and restoring backgrounds in experimental tests.