MambaVSR: Content-Aware Scanning State Space Model for Video Super-Resolution

TL;DR

MambaVSR introduces a novel content-aware state-space model for video super-resolution that effectively models non-local dependencies and improves detail recovery with fewer parameters.

Contribution

It is the first to incorporate a content-aware scanning mechanism within a state-space model framework for VSR, enhancing non-local dependency modeling and efficiency.

Findings

Outperforms Transformer-based methods by 0.58 dB PSNR on REDS dataset.

Uses 55% fewer parameters than existing methods.

Effectively models non-local dependencies and preserves details.

Abstract

Video super-resolution (VSR) faces critical challenges in effectively modeling non-local dependencies across misaligned frames while preserving computational efficiency. Existing VSR methods typically rely on optical flow strategies or transformer architectures, which struggle with large motion displacements and long video sequences. To address this, we propose MambaVSR, the first state-space model framework for VSR that incorporates an innovative content-aware scanning mechanism. Unlike rigid 1D sequential processing in conventional vision Mamba methods, our MambaVSR enables dynamic spatiotemporal interactions through the Shared Compass Construction (SCC) and the Content-Aware Sequentialization (CAS). Specifically, the SCC module constructs intra-frame semantic connectivity graphs via efficient sparse attention and generates adaptive spatial scanning sequences through spectral clustering. Building upon SCC, the CAS module effectively aligns and aggregates non-local similar content across multiple frames by interleaving temporal features along the learned spatial order. To bridge global dependencies with local details, the Global-Local State Space Block (GLSSB) synergistically integrates window self-attention operations with SSM-based feature propagation, enabling high-frequency detail recovery under global dependency guidance. Extensive experiments validate MambaVSR's superiority, outperforming the Transformer-based method by 0.58 dB PSNR on the REDS dataset with 55% fewer parameters.