Meta-Transfer Learning for Zero-Shot Super-Resolution

By Jae Woong Soh, Sunwoo Cho, Nam Ik Cho. Computer Vision and Pattern Recognition, 2020

320 citations recorded by the source corpus.

Convolutional neural networks (CNNs) have shown dramatic improvements in single image super-resolution (SISR) by using large-scale external samples. Despite their remarkable performance based on the external dataset, they cannot exploit internal information within a specific image. Another problem is that they are applicable only to the specific condition of data that they are supervised. For instance, the low-resolution (LR) image should be a "bicubic" downsampled noise-free image from a high-resolution (HR) one. To address both issues, zero-shot super-resolution (ZSSR) has been proposed for flexible internal learning. However, they require thousands of gradient updates, i.e., long inference time. In this paper, we present Meta-Transfer Learning for Zero-Shot Super-Resolution (MZSR), which leverages ZSSR. Precisely, it is based on finding a generic initial parameter that is suitable for internal learning. Thus, we can exploit both external and internal information, where one single gradient update can yield quite considerable results. With our method, the network can quickly adapt to a given image condition. In this respect, our method can be applied to a large spectrum of image conditions within a fast adaptation process.

View Paper: https://www.semanticscholar.org/paper/385931984baef107aa9ee8ba28da1463f0beba36

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Real-World Single Image Super-Resolution: A Brief Review (2021, methodology): “[114] present the meta-transfer learning for zero-shot SR (MZSR), which consists of three steps, i.”
Flow-based Kernel Prior with Application to Blind Super-Resolution (2021, background): “Since the blur kernels of real LR images are usually unimodal [35, 41] and can typically be modeled by a Gaussian [40], most existing blind SR works [3, 17, 21, 40, 42, 43, 49, 54, 55] assume the SR kernel is isotropic or anisotropic Gaussian kernel.”
Unsupervised Degradation Representation Learning for Blind Super-Resolution (2021, methodology): “To super-resolve real images with unknown degradations, degradation estimation [28, 3] needs to be performed to provide degradation information for non-blind SR networks [42, 33, 40, 34].”
Plug-and-Play Image Restoration With Deep Denoiser Prior (2020, methodology): “For the compared methods, we consider the bicubic interpolation method, RCAN [84], SRFBN [85], MZSR [86], IRCNN and IRCNN+. Speciﬁcally, RCAN is the state-of-the-art bicubic degradation based deep model consisting of about 400 layers.”
LARGE: Latent-Based Regression through GAN Semantics (2021, background): “They tackle image-to-image translation [31], super resolution [56], motion prediction [17], and re-identiﬁcation [68].”