Efficient neural supersampling on a novel gaming dataset

TL;DR

This paper presents a highly efficient neural supersampling algorithm for real-time gaming rendering, supported by a new dataset with auxiliary modalities to facilitate progress in super-resolution techniques.

Contribution

Introduces a novel neural supersampling method that is four times more efficient while maintaining accuracy, and provides a new dataset with auxiliary modalities for benchmarking.

Findings

Neural supersampling is significantly more efficient.

The new dataset enhances benchmarking capabilities.

Maintains accuracy comparable to existing methods.

Abstract

Real-time rendering for video games has become increasingly challenging due to the need for higher resolutions, framerates and photorealism. Supersampling has emerged as an effective solution to address this challenge. Our work introduces a novel neural algorithm for supersampling rendered content that is 4 times more efficient than existing methods while maintaining the same level of accuracy. Additionally, we introduce a new dataset which provides auxiliary modalities such as motion vectors and depth generated using graphics rendering features like viewport jittering and mipmap biasing at different resolutions. We believe that this dataset fills a gap in the current dataset landscape and can serve as a valuable resource to help measure progress in the field and advance the state-of-the-art in super-resolution techniques for gaming content.