Brick Pose Estimation on BC-Bench Best Object Performance
14.29Translation Error (LDU)Brick-Composer
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Brick-ComposerBackbone=Qwen-3-8B-VL, Evaluation Protocol=Integrated learning2026.06 | 14.29 | 0 | |
| World Feedback (L)Backbone=Qwen-3-8B-VL, Evaluation Protocol=Simulator-generated feedback2026.06 | 21.46 | 11.64 | |
| Brick-ComposerBackbone=Gemma-3-12B, Evaluation Protocol=Integrated learning2026.06 | 24.63 | 12.86 | |
| Designer SupervisionBackbone=Qwen-3-8B-VL, Evaluation Protocol=Human supervision2026.06 | 27.13 | 12.95 | |
| World Feedback (L)Backbone=Gemma-3-12B, Evaluation Protocol=Simulator-generated feedback2026.06 | 27.56 | 12.86 | |
| Designer SupervisionBackbone=Gemma-3-12B, Evaluation Protocol=Human supervision2026.06 | 36.91 | 15 | |
| World Feedback (P)Backbone=Gemma-3-12B, Evaluation Protocol=Simulator feedback2026.06 | 39.46 | 14.53 | |
| InternVL-3.5-8BEvaluation Protocol=Zero-shot2026.06 | 41.54 | 35 | |
| Gemma-3-12BEvaluation Protocol=Zero-shot2026.06 | 41.86 | 12.86 | |
| World Feedback (P)Backbone=Qwen-3-8B-VL, Evaluation Protocol=Simulator feedback2026.06 | 42.36 | 12.86 | |
| Qwen-3-VL-8BEvaluation Protocol=Zero-shot2026.06 | 43.81 | 12.86 | |
| Qwen-3.5-VL-27BEvaluation Protocol=Zero-shot2026.06 | 64.47 | 34.84 | |
| GPT-5.4Evaluation Protocol=Zero-shot2026.06 | 67.94 | 40 |