Recursive Neural Network for Video Deblurring

• Published in: IEEE transactions on circuits and systems for video technology Vol. 31; no. 8; pp. 3025 - 3036
• Main Authors: Zhang, Xiaoqin, Jiang, Runhua, Wang, Tao, Wang, Jinxin
• Format: Journal Article
• Language: English
• Published: New York IEEE 01.08.2021; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Computational modeling; Convolution; Feature maps; Image restoration; Iterative methods; Kernel; Neural networks; Recurrent neural networks; recursive neural network; Task analysis; temporal consistency; Video deblurring
• ISSN: 1051-8215; 1558-2205
• DOI: 10.1109/TCSVT.2020.3035722

Video deblurring is still a challenging low-level vision task since spatio-temporal characteristics across both the spatial and temporal domains are difficult to model. In this article, to model the temporal information, we develop a non-local block which estimates inter-frame similarity and inter-frame difference. Specially, for modeling the spatial characteristics and restoring sharp frame details, we propose a recursive block that iteratively refines feature maps generated at the last iteration. In addition, a novel temporal loss function is introduced to ensure the temporal consistency of generated frames. Experimental results on public datasets demonstrate that our method achieves state-of-the-art performance both quantitatively and qualitatively.