MatrixCity: A Large-scale City Dataset for City-scale Neural Rendering and Beyond

TL;DR

MatrixCity is a large-scale synthetic city dataset designed to advance neural rendering at city scale, providing diverse views, ground-truth data, and environmental controls to facilitate research and address current challenges.

Contribution

The paper introduces MatrixCity, a comprehensive synthetic dataset for city-scale neural rendering, including a pipeline for data collection and a benchmark highlighting unique challenges.

Findings

MatrixCity contains 67k aerial and 452k street images.

Benchmark reveals specific challenges in city-scale neural rendering.

Dataset supports various environmental conditions for diverse research needs.

Abstract

Neural radiance fields (NeRF) and its subsequent variants have led to remarkable progress in neural rendering. While most of recent neural rendering works focus on objects and small-scale scenes, developing neural rendering methods for city-scale scenes is of great potential in many real-world applications. However, this line of research is impeded by the absence of a comprehensive and high-quality dataset, yet collecting such a dataset over real city-scale scenes is costly, sensitive, and technically difficult. To this end, we build a large-scale, comprehensive, and high-quality synthetic dataset for city-scale neural rendering researches. Leveraging the Unreal Engine 5 City Sample project, we develop a pipeline to easily collect aerial and street city views, accompanied by ground-truth camera poses and a range of additional data modalities. Flexible controls over environmental factors like light, weather, human and car crowd are also available in our pipeline, supporting the need of various tasks covering city-scale neural rendering and beyond. The resulting pilot dataset, MatrixCity, contains 67k aerial images and 452k street images from two city maps of total size $28km^2$. On top of MatrixCity, a thorough benchmark is also conducted, which not only reveals unique challenges of the task of city-scale neural rendering, but also highlights potential improvements for future works. The dataset and code will be publicly available at our project page: https://city-super.github.io/matrixcity/.