May 10, 2022 · We propose a metric learning strategy to map the unquantifiable degradation levels in real-world scenarios to a metric space, which is trained in an ...

In this paper, we present the first attempt to study the interactive modulation in real-world super-resolution through a novel metric learning based strategy.

Metric Learning based Interactive Modulation for Real-World Super-Resolution. arXiv Chong Mou, Yanze Wu, Xintao Wang, Chao Dong, Jian Zhang, Ying Shan

Oct 24, 2022 · We propose a metric learning strategy to map the unquantifiable degradation levels in real-world scenarios to a metric space, which is trained in an ...

A metric learning strategy to map the unquantifiable degradation levels in real-world scenarios to a metric space, which is trained in an unsupervised ...

Empowering Real-World Image Super-Resolution with Flexible Interactive Modulation · ReVideo: Remake a Video with Motion and Content Control · DiffEditor: Boosting ...

Extensive experiments demonstrate that the proposed MM-RealSR achieves excellent modulation and restoration performance in real-world super-resolution. Codes ...

There are two main differences between our proposed MM-RealSR and existing modulation-based methods. First, existing methods are limited to be trained.

Interactive image restoration aims to restore images by adjusting several controlling coefficients, which determine the restoration strength.

Overview of our proposed metric-learning-based interactive modulation method for real-world super-resolution (MM-RealSR+). It is composed of a base network, a ...