3D Room Generation from Top-down Image on Code-as-Room (CaR) 1.0 (test)
67.5Object RecallGPT-5.5 w/CaR
Evaluation Results
| Method | Links | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| GPT-5.5 w/CaRVLM Backbone=GPT-5.5, Workflow=Code-as-Room (CaR)2026.05 | 67.5 | 72.54 | 10.5 | 66.7 | 71.4 | 92.2 | 80.1 | 71.1 | 73.3 | 7.28 | 6 | 7.52 | |
| Gemini3-flash w/CaRVLM Backbone=Gemini-3 Flash, Workflow=Code-as-Room (CaR)2026.05 | 58.9 | 88.42 | 2.57 | 72 | 76.9 | 93.5 | 93.5 | 100 | 100 | 8.32 | 6.07 | 7.17 | |
| Gemini3.1-pro w/CaRVLM Backbone=Gemini-3.1 Pro, Workflow=Code-as-Room (CaR)2026.05 | 55.5 | 84.3 | 3.3 | 73.2 | 79.8 | 93.6 | 94 | 100 | 95.5 | 8.08 | 7.05 | 8.2 | |
| GPT-5.5VLM Backbone=GPT-5.5, Workflow=Direct Generation2026.05 | 42.2 | 71.7 | 14.5 | 46.2 | 50.8 | 65.3 | 52.6 | — | 42.2 | 5.8 | 5.42 | 7.24 | |
| Gemini3.1-proVLM Backbone=Gemini-3.1 Pro, Workflow=Direct Generation2026.05 | 17.8 | 15.3 | 8.4 | 16.8 | 54.7 | 78 | 41.7 | — | 57.8 | 2.49 | 3.24 | 2.51 |