Stable Long-Term Recurrent Video Super-Resolution

TL;DR

This paper identifies instability issues in recurrent video super-resolution models on long, low-motion sequences and proposes a new stable, competitive framework and network architecture based on Lipschitz stability theory.

Contribution

It introduces a novel stability framework for recurrent VSR models, a new dataset for long sequences, and a stable, high-performing VSR network called MRVSR.

Findings

Recurrent VSR models diverge on low-motion long sequences.

The proposed MRVSR network maintains stability and performance.

Empirical results show MRVSR's competitiveness on the new dataset.

Abstract

Recurrent models have gained popularity in deep learning (DL) based video super-resolution (VSR), due to their increased computational efficiency, temporal receptive field and temporal consistency compared to sliding-window based models. However, when inferring on long video sequences presenting low motion (i.e. in which some parts of the scene barely move), recurrent models diverge through recurrent processing, generating high frequency artifacts. To the best of our knowledge, no study about VSR pointed out this instability problem, which can be critical for some real-world applications. Video surveillance is a typical example where such artifacts would occur, as both the camera and the scene stay static for a long time. In this work, we expose instabilities of existing recurrent VSR networks on long sequences with low motion. We demonstrate it on a new long sequence dataset Quasi-Static Video Set, that we have created. Finally, we introduce a new framework of recurrent VSR networks that is both stable and competitive, based on Lipschitz stability theory. We propose a new recurrent VSR network, coined Middle Recurrent Video Super-Resolution (MRVSR), based on this framework. We empirically show its competitive performance on long sequences with low motion.