InternScenes: A Large-scale Simulatable Indoor Scene Dataset with Realistic Layouts

2025-09-13

AI Summary

Existing large-scale indoor 3D scene datasets suffer from limited scale, oversimplified layouts, scarcity of small objects, and severe inter-object collisions. To address these limitations, this work introduces a large-scale, simulation-ready dataset comprising approximately 40,000 diverse scenes spanning 15 scene categories and 288 object classes, synthesized from three complementary sources: real-world scans, procedural generation, and manual design. We propose a multi-source scene fusion framework that preserves fine-grained small objects and resolves layout conflicts via physics-based collision elimination. The resulting scenes are rendered as high-fidelity, interactive simulation replicas. Extensive evaluation on scene layout generation and point-goal navigation demonstrates significant improvements in model generalization and reasoning capabilities within complex, realistic environments. This dataset establishes a high-quality foundational resource for large-scale training of embodied AI systems.

Technology Category

• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Search and Optimization: Sampling/Simulation-based Search
• Computer Vision: 3D Computer Vision

Application Category

• Web Mining and Content Analysis: Web data generation and simulation
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies

Abstract

The advancement of Embodied AI heavily relies on large-scale, simulatable 3D scene datasets characterized by scene diversity and realistic layouts. However, existing datasets typically suffer from limitations in data scale or diversity, sanitized layouts lacking small items, and severe object collisions. To address these shortcomings, we introduce extbf{InternScenes}, a novel large-scale simulatable indoor scene dataset comprising approximately 40,000 diverse scenes by integrating three disparate scene sources, real-world scans, procedurally generated scenes, and designer-created scenes, including 1.96M 3D objects and covering 15 common scene types and 288 object classes. We particularly preserve massive small items in the scenes, resulting in realistic and complex layouts with an average of 41.5 objects per region. Our comprehensive data processing pipeline ensures simulatability by creating real-to-sim replicas for real-world scans, enhances interactivity by incorporating interactive objects into these scenes, and resolves object collisions by physical simulations. We demonstrate the value of InternScenes with two benchmark applications: scene layout generation and point-goal navigation. Both show the new challenges posed by the complex and realistic layouts. More importantly, InternScenes paves the way for scaling up the model training for both tasks, making the generation and navigation in such complex scenes possible. We commit to open-sourcing the data, models, and benchmarks to benefit the whole community.

Problem

• Creating large-scale simulatable indoor scenes with realistic layouts
• Addressing limitations in scene diversity and object collisions
• Integrating multiple scene sources for enhanced complexity and interactivity

Innovation

• Integrates three disparate scene sources
• Ensures simulatability via real-to-sim replicas
• Resolves object collisions through physical simulations