PulseAugur
LIVE 09:44:36
research · [1 source] ·
0
research

InternScenes dataset offers large-scale, realistic indoor environments for Embodied AI

Researchers have introduced InternScenes, a large-scale dataset designed to advance Embodied AI research. This dataset features approximately 40,000 diverse indoor scenes, integrating real-world scans, procedural generation, and designer-created environments. InternScenes aims to overcome limitations of existing datasets by including a vast number of small items and realistic, complex layouts with an average of 41.5 objects per region, while also ensuring simulatability and interactivity. The dataset's utility is demonstrated through benchmarks in scene layout generation and point-goal navigation, with plans to open-source the data, models, and benchmarks. AI

Summary written by gemini-2.5-flash-lite from 1 source. How we write summaries →

IMPACT Provides a large-scale, realistic dataset to improve training for Embodied AI tasks like navigation and scene generation.

RANK_REASON This is a research paper introducing a new dataset for Embodied AI.

Read on arXiv cs.CV →

COVERAGE [1]

  1. arXiv cs.CV TIER_1 · Weipeng Zhong, Peizhou Cao, Yichen Jin, Li Luo, Wenzhe Cai, Jingli Lin, Hanqing Wang, Zhaoyang Lyu, Tai Wang, Bo Dai, Xudong Xu, Jiangmiao Pang ·

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

    arXiv:2509.10813v4 Announce Type: replace 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 di…