InternScenes is a large-scale simulatable indoor scene dataset with realistic layouts, released in 2025 by Shanghai Artificial Intelligence Laboratory, Shanghai Jiao Tong University, Beihang University and other institutions. The related paper results are "InternScenes: A Large-scale Simulatable Indoor Scene Dataset with Realistic Layouts".

This dataset consists of approximately 40,000 diverse scenes and 1.96 million 3D objects, covering 15 typical indoor scene types and 288 object categories. Its overall size is approximately 10 times that of existing similar datasets. Compared to previous datasets, this dataset specifically retains a large number of small objects, thereby constructing more realistic and complex scene layouts. Each area contains an average of 41.5 objects. Approximately 20% of these objects are interactive objects, covering common everyday items such as cabinets, microwave ovens, ovens, and refrigerators.