For more information and statistics please see our paper.

Distribution of physical sizes

All object meshes and 3D scenes in the SUNCG dataset are measured in real-world spatial dimensions (units are in meters). This plot shows statistics related to physical size over three levels: rooms, floors and houses.

Distribution of different room types

The four most common room types in our dataset are bedroom, living room, kitchen and toilet, which agrees with the distribution in real-world living spaces.

Object-Room Relationship

Left shows the distribution of number of objects in each room. On average there are more than 14 objects in each room. On the right we show the object category distribution conditioned on different room types. Size of the square shows the frequency of given object category appears in the certain room type. The frequency is normalized for each object category. As expected, object occurrences are tightly correlated with the room type. For example, kitchen counter has a very high chance to appear in kitchen, tvs are more likely to appear in living room or bedroom, while chairs appear frequently in many room types.

Scene structure statistics.

The 3D scenes in our dataset are range from single room studio to multi-floor houses. The average and median number of rooms per-house are 8.9 and 7 respectively. The average and median number of floors per-house are 1.3 and 1 respectively.