Adapting Image Super-Resolution State-of-the-arts and Learning Multi-model Ensemble for Video Super-Resolution

TL;DR

This paper adapts advanced image super-resolution techniques for video super-resolution and introduces a learning-based ensemble method, achieving high performance in a competitive benchmark.

Contribution

It presents a simple adaptation approach for applying image SR methods to videos and proposes a multi-model ensemble strategy for improved results.

Findings

Achieved second place in NTIRE2019 Video SR Challenge.

Effective exploitation of temporal information with minimal overhead.

Enhanced super-resolution performance through model ensembling.

Abstract

Recently, image super-resolution has been widely studied and achieved significant progress by leveraging the power of deep convolutional neural networks. However, there has been limited advancement in video super-resolution (VSR) due to the complex temporal patterns in videos. In this paper, we investigate how to adapt state-of-the-art methods of image super-resolution for video super-resolution. The proposed adapting method is straightforward. The information among successive frames is well exploited, while the overhead on the original image super-resolution method is negligible. Furthermore, we propose a learning-based method to ensemble the outputs from multiple super-resolution models. Our methods show superior performance and rank second in the NTIRE2019 Video Super-Resolution Challenge Track 1.