Low-Resource Video Super-Resolution using Memory, Wavelets, and Deformable Convolutions

TL;DR

This paper introduces a lightweight, parameter-efficient neural network for video super-resolution that combines wavelet decompositions, a single memory tensor, and deformable convolutions to achieve high-quality results suitable for edge devices.

Contribution

The paper presents a novel VSR architecture that is both lightweight and highly accurate, utilizing wavelet transforms, a single memory tensor, and residual deformable convolutions for improved efficiency.

Findings

Achieves state-of-the-art accuracy with only 2.3 million parameters.

Utilizes wavelet decompositions for spatial sparsity prior.

Employs a single memory tensor for efficient inter-frame information capture.

Abstract

The tradeoff between reconstruction quality and compute required for video super-resolution (VSR) remains a formidable challenge in its adoption for deployment on resource-constrained edge devices. While transformer-based VSR models have set new benchmarks for reconstruction quality in recent years, these require substantial computational resources. On the other hand, lightweight models that have been introduced even recently struggle to deliver state-of-the-art reconstruction. We propose a novel lightweight and parameter-efficient neural architecture for VSR that achieves state-of-the-art reconstruction accuracy with just 2.3 million parameters. Our model enhances information utilization based on several architectural attributes. Firstly, it uses 2D wavelet decompositions strategically interlayered with learnable convolutional layers to utilize the inductive prior of spatial sparsity of edges in visual data. Secondly, it uses a single memory tensor to capture inter-frame temporal information while avoiding the computational cost of previous memory-based schemes. Thirdly, it uses residual deformable convolutions for implicit inter-frame object alignment that improve upon deformable convolutions by enhancing spatial information in inter-frame feature differences. Architectural insights from our model can pave the way for real-time VSR on the edge, such as display devices for streaming data.