| Paper | Code | GPT-4 | Sentence-BERT | SimCSE | Precision | Correctness | Hallucination | ModelName | ReleaseDate |
|---|---|---|---|---|---|---|---|---|---|
| MiniGPT-3D: Efficiently Aligning 3D Point Clouds with Large Language Models using 2D Priors | ✓ Link | 57.06 | 49.54 | 51.39 | 83.14 | 3.50 | 0.71 | MiniGPT-3D | 2024-05-02 |
| ShapeLLM: Universal 3D Object Understanding for Embodied Interaction | ✓ Link | 48.94 | 48.52 | 49.98 | ShapeLLM-13B | 2024-02-27 | |||
| PointLLM: Empowering Large Language Models to Understand Point Clouds | ✓ Link | 48.15 | 47.91 | 49.12 | 78.75 | 3.10 | 0.84 | PointLLM-13B V1.2 | 2023-08-31 |
| ShapeLLM: Universal 3D Object Understanding for Embodied Interaction | ✓ Link | 46.92 | 48.20 | 49.23 | ShapeLLM-7B | 2024-02-27 | |||
| PointLLM: Empowering Large Language Models to Understand Point Clouds | ✓ Link | 44.85 | 47.47 | 48.55 | 82.14 | 3.04 | 0.66 | PointLLM-7B V1.2 | 2023-08-31 |
| 3D-LLM: Injecting the 3D World into Large Language Models | ✓ Link | 33.42 | 44.48 | 43.68 | 60.39 | 1.77 | 1.16 | 3D-LLM | 2023-07-24 |