Boosting Neural Representations for Videos with a Conditional Decoder

TL;DR

This paper presents a universal boosting framework for implicit neural video representations, improving their reconstruction quality, convergence speed, and codec performance through a conditional decoder and novel feature generation methods.

Contribution

It introduces a conditional decoder with a temporal-aware affine transform and sinusoidal NeRV-like blocks to enhance implicit video representations, a novel approach not previously explored.

Findings

Boosts baseline INRs' reconstruction quality and convergence speed.

Achieves superior inpainting and interpolation results.

Outperforms baseline INRs and rivals traditional codecs in rate-distortion performance.

Abstract

Implicit neural representations (INRs) have emerged as a promising approach for video storage and processing, showing remarkable versatility across various video tasks. However, existing methods often fail to fully leverage their representation capabilities, primarily due to inadequate alignment of intermediate features during target frame decoding. This paper introduces a universal boosting framework for current implicit video representation approaches. Specifically, we utilize a conditional decoder with a temporal-aware affine transform module, which uses the frame index as a prior condition to effectively align intermediate features with target frames. Besides, we introduce a sinusoidal NeRV-like block to generate diverse intermediate features and achieve a more balanced parameter distribution, thereby enhancing the model's capacity. With a high-frequency information-preserving reconstruction loss, our approach successfully boosts multiple baseline INRs in the reconstruction quality and convergence speed for video regression, and exhibits superior inpainting and interpolation results. Further, we integrate a consistent entropy minimization technique and develop video codecs based on these boosted INRs. Experiments on the UVG dataset confirm that our enhanced codecs significantly outperform baseline INRs and offer competitive rate-distortion performance compared to traditional and learning-based codecs. Code is available at https://github.com/Xinjie-Q/Boosting-NeRV.