# GeoProg3D: Compositional Visual Reasoning for City-Scale 3D Language Fields

arXiv: https://arxiv.org/abs/2506.23352 · HTML: https://mlsanity.com/paper/2506.23352/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Placed in Spatial and embodied reasoning, one of 17 directions in LLM reasoning.
- Builds on 14 of its 95 references that are in the map.
- First posted on arXiv 29 Jun 2025.

## Placed in

- LLM reasoning / [Spatial and embodied reasoning](https://mlsanity.com/lr/spatial-and-embodied-reasoning/)

## Builds on (most cited first, 14 of 14)

- [Qwen2.5-VL Technical Report](https://mlsanity.com/paper/2502.13923/) (2025)
- [LayoutGPT: Compositional Visual Planning and Generation with Large Language Models](https://mlsanity.com/paper/2305.15393/) (2023)
- [RSGPT: A Remote Sensing Vision Language Model and Benchmark](https://mlsanity.com/paper/2307.15266/) (2023)
- [SatCLIP: Global, General-Purpose Location Embeddings with Satellite Imagery](https://mlsanity.com/paper/2311.17179/) (2023)
- [StreetSurf: Extending Multi-view Implicit Surface Reconstruction to Street Views](https://mlsanity.com/paper/2306.04988/) (2023)
- [SkySenseGPT: A Fine-Grained Instruction Tuning Dataset and Model for Remote Sensing Vision-Language Understanding](https://mlsanity.com/paper/2406.10100/) (2024)
- [TEOChat: A Large Vision-Language Assistant for Temporal Earth Observation Data](https://mlsanity.com/paper/2410.06234/) (2024)
- [MiniGPT-3D: Efficiently Aligning 3D Point Clouds with Large Language Models using 2D Priors](https://mlsanity.com/paper/2405.01413/) (2024)
- [Generative Gaussian Splatting for Unbounded 3D City Generation](https://mlsanity.com/paper/2406.06526/) (2024)
- [GraspSplats: Efficient Manipulation with 3D Feature Splatting](https://mlsanity.com/paper/2409.02084/) (2024)
- [Efficient Large-scale Scene Representation with a Hybrid of High-resolution Grid and Plane Features](https://mlsanity.com/paper/2303.03003/) (2023)
- [Towards Vision-Language Geo-Foundation Model: A Survey](https://mlsanity.com/paper/2406.09385/) (2024)
- [EarthMarker: A Visual Prompting Multi-modal Large Language Model for Remote Sensing](https://mlsanity.com/paper/2407.13596/) (2024)
- [PreSight: Enhancing Autonomous Vehicle Perception with City-Scale NeRF Priors](https://mlsanity.com/paper/2403.09079/) (2024)

## Built on by (most cited first, 3 of 3)

- [The Semantic Lifecycle in Embodied AI: Acquisition, Representation and Storage via Foundation Models](https://mlsanity.com/paper/2601.08876/) (2026)
- [3DCity-LLM: Empowering Multi-modality Large Language Models for 3D City-scale Perception and Understanding](https://mlsanity.com/paper/2603.23447/) (2026)
- [Stitch4D: Sparse Multi-Location 4D Urban Reconstruction via Spatio-Temporal Interpolation](https://mlsanity.com/paper/2604.07923/) (2026)

## Abstract (arXiv)

The advancement of 3D language fields has enabled intuitive interactions with 3D scenes via natural language. However, existing approaches are typically limited to small-scale environments, lacking the scalability and compositional reasoning capabilities necessary for large, complex urban settings. To overcome these limitations, we propose GeoProg3D, a visual programming framework that enables natural language-driven interactions with city-scale high-fidelity 3D scenes. GeoProg3D consists of two key components: (i) a Geography-aware City-scale 3D Language Field (GCLF) that leverages a memory-efficient hierarchical 3D model to handle large-scale data, integrated with geographic information for efficiently filtering vast urban spaces using directional cues, distance measurements, elevation data, and landmark references; and (ii) Geographical Vision APIs (GV-APIs), specialized geographic vision tools such as area segmentation and object detection. Our framework employs large language models (LLMs) as reasoning engines to dynamically combine GV-APIs and operate GCLF, effectively supporting diverse geographic vision tasks. To assess performance in city-scale reasoning, we introduce GeoEval3D, a comprehensive benchmark dataset containing 952 query-answer pairs across five challenging tasks: grounding, spatial reasoning, comparison, counting, and measurement. Experiments demonstrate that GeoProg3D significantly outperforms existing 3D language fields and vision-language models across multiple tasks. To our knowledge, GeoProg3D is the first framework enabling compositional geographic reasoning in high-fidelity city-scale 3D environments via natural language. The code is available at https://snskysk.github.io/GeoProg3D/.