MRFN: Multi-Receptive-Field Network for Fast and Accurate Single Image Super-Resolution

• Published in: IEEE transactions on multimedia Vol. 22; no. 4; pp. 1042 - 1054
• Main Authors: He, Zewei, Cao, Yanpeng, Du, Lei, Xu, Baobei, Yang, Jiangxin, Cao, Yanlong, Tang, Siliang, Zhuang, Yueting
• Format: Journal Article
• Language: English
• Published: Piscataway IEEE 01.04.2020; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Artificial neural networks; Back propagation; Computational modeling; Convolutional neural networks; deep learning; Feature extraction; Image reconstruction; Image resolution; Image restoration; loss function; Modules; multi-receptive-field; Restoration; Single image super-resolution; Training
• ISSN: 1520-9210; 1941-0077
• DOI: 10.1109/TMM.2019.2937688

Recently, convolutional neural network (CNN) based models have shown great potential in the task of single image super-resolution (SISR). However, many state-of-the-art SISR solutions are reproducing some tricks proven effective in other vision tasks, such as pursuing a deeper model. In this paper, we propose a new solution (named as Multi-Receptive-Field Network - MRFN), which outperforms existing SISR solutions in three different aspects. First, from receptive field: a novel multi-receptive-field (MRF) module is proposed to extract and fuse features in different receptive fields from local to global. Integrating these hierarchical features can generate better mappings on recovering high-fidelity details at different scales. Second, from network architectures: both dense skip connections and deep supervision are utilized to combine features from the current MRF module and preceding ones for training more representative features. Moreover, a deconvolution layer is embedded at the end of the network to avoid artificial priors induced by numerical data pre-processing (e.g., bicubic stretching), and speed up the restoration process. Finally, from error modeling: different from <inline-formula><tex-math notation="LaTeX">L1</tex-math></inline-formula> and <inline-formula><tex-math notation="LaTeX">L2</tex-math></inline-formula> loss functions, we proposed a novel two-parameter training loss called Weighted Huber loss function which can adaptively adjust the value of back-propagated derivative according to the residual value, thus fit the reconstruction error more effectively. Extensive qualitative and quantitative evaluation results on benchmark datasets demonstrate that our proposed MRFN can achieve more accurate recovering results than most state-of-the-art methods with significantly less complexity.