Fig 1.
Model structure (The left is a Transformer model; The right is a diffusion model).
Fig 2.
The algorithm for the Transformer-DN model.
Table 1.
Dataset statistics.
Table 2.
Hardware information.
Table 3.
Model parameters.
Fig 3.
Comparison of design times for models in the InteriorNet dataset.
Fig 4.
Comparison of model design times in the SUN RGB-D dataset.
Fig 5.
Comparison of design times for models in the NYU Depth V2 dataset.
Fig 6.
Comparison of model design times in the ScanNet dataset.
Fig 7.
Evaluation results of space utilization of models (a: The InteriorNet dataset; b: The SUN RGB-D dataset).
Fig 8.
Evaluation results of space utilization of models (a: The NYU Depth V2 dataset; b: The ScanNet dataset).
Table 4.
Enhancement effect of model proportional and scale coordination.
Table 5.
Assessment results of color personalization of the model.
Table 6.
Results of significance analysis.