WildCity: A Real-World City-Scale Testbed for Rendering, Simulation, and Spatial Intelligence

Hugging Face Daily Papers Papers
city-scale multimodal-dataset spatial-intelligence urban-navigation scene-reconstruction simulation benchmark

Summary

WildCity introduces a large-scale multimodal dataset for city-scale urban navigation and spatial representation, collected by autonomous fleets. It provides 18 long trajectories and establishes baselines for reconstruction and closed-loop simulation to advance AI systems that can perceive and reason about city-scale environments.

Humans can navigate an unfamiliar city and gradually form a coherent spatial mental map spanning tens of square kilometers. Can AI build spatial representations at a comparable scale? Although recent foundation models have advanced scene reconstruction and embodied intelligence, scaling to entire cities remains an open challenge, primarily due to the lack of city-scale data. To bridge the gap, we introduce WildCity, a real-world multimodal dataset collected by autonomous fleets traversing complex urban environments. Our dataset includes 18 trajectories, each averaging 83.7 kilometers in length, and preserves the core challenges of in-the-wild perception, e.g., dynamic objects, lighting variations, and imperfect camera poses. We further establish an urban-tailored reconstruction baseline and convert the reconstructed environments into a closed-loop simulator. Beyond the dataset and baseline, we systematically analyze the key challenges on the path to simulation-ready urban digital twins: scalability, extrapolation, and uncertainty. Ultimately, WildCity aims to catalyze progress not only in city-scale rendering, but more broadly in the pursuit of AI that can perceive, remember, and reason across space at a scale comparable to human cognition. Project page: https://han-xiangyu.github.io/Wild-City/
Original Article
View Cached Full Text

Cached at: 07/09/26, 07:51 AM

Paper page - WildCity: A Real-World City-Scale Testbed for Rendering, Simulation, and Spatial Intelligence

Source: https://huggingface.co/papers/2607.06838 Authors:

,

,

,

,

,

,

,

,

,

,

Abstract

WildCity presents a large-scale multimodal dataset for urban navigation and spatial representation, enabling research into AI systems that can perceive and reason about city-scale environments similar to human cognitive capabilities.

Humans can navigate an unfamiliar city and gradually form a coherent spatial mental map spanning tens of square kilometers. Can AI buildspatial representations at a comparable scale? Although recent foundation models have advancedscene reconstructionandembodied intelligence, scaling to entire cities remains an open challenge, primarily due to the lack ofcity-scale data. To bridge the gap, we introduce WildCity, a real-worldmultimodal datasetcollected byautonomous fleetstraversing complexurban environments. Our dataset includes 18 trajectories, each averaging 83.7 kilometers in length, and preserves the core challenges of in-the-wildperception, e.g., dynamic objects, lighting variations, and imperfect camera poses. We further establish an urban-tailored reconstruction baseline and convert the reconstructed environments into aclosed-loop simulator. Beyond the dataset and baseline, we systematically analyze the key challenges on the path to simulation-readyurban digital twins: scalability, extrapolation, and uncertainty. Ultimately, WildCity aims to catalyze progress not only in city-scale rendering, but more broadly in the pursuit of AI that can perceive, remember, and reason across space at a scale comparable to human cognition. Project page: https://han-xiangyu.github.io/Wild-City/

View arXiv pageView PDFProject pageGitHub2Add to collection

Get this paper in your agent:

hf papers read 2607\.06838

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2607.06838 in a model README.md to link it from this page.

Datasets citing this paper1

#### Neptune615/Wild-City Preview• Updatedabout 2 hours ago • 24 • 1

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2607.06838 in a Space README.md to link it from this page.

Collections including this paper1

Similar Articles

360CityArena: A Realistic Virtual Urban Navigation Benchmark for Embodied Agents

Hugging Face Daily Papers

360CityArena is a photorealistic urban navigation benchmark built from 360-degree videos of Tokyo's Akihabara district. It evaluates embodied agents' environment understanding, path reasoning, and spatial reasoning, showing that even strong LMM-based agents like Gemini 2.5 Flash perform far below human level.