3D Object Classification on ModelNet40 (test) (Prompt-based Evaluation)
62.4Accuracy (I prompt)PointLLM-R-7B
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| PointLLM-R-7Bzero-shot=true, parameters=7B, protocol=PointLLM2026.05 | 62.4 | 61.6 | — | |
| MiniGPT-3DLLM Size=2.7B, 3D Data Size=730K, Input=Point Cloud2026.02 | 61.99 | 60.49 | 61.24 | |
| MiniGPT-3DLLM Size=2.7B, Trainable Params=0.05B (47.8M), Input=3D Point Cloud2024.05 | 61.75 | 59.97 | 60.86 | |
| PointAlignLLM Size=2.7B, 3D Data Size=730K, Input=Point Cloud2026.02 | 61.55 | 60.78 | 61.17 | |
| MiniGPT-3Dzero-shot=true, protocol=PointLLM2026.05 | 57.49 | 57.63 | — | |
| PointLLM-13Bzero-shot=true, parameters=13B, protocol=PointLLM2026.05 | 53.79 | 53.99 | — | |
| PointLLM-7BLLM Size=7B, Trainable Params=7.01B, Input=3D Point Cloud2024.05 | 53.44 | 51.82 | 52.63 | |
| PointLLM-7BInput=3D Point Cloud, Zero-shot=true2023.08 | 53.44 | 51.82 | 52.82 | |
| PointLLM-7Bzero-shot=true, parameters=7B, protocol=PointLLM2026.05 | 53.24 | 53.42 | — | |
| PointLLM-13BLLM Size=13B, Trainable Params=13.01B, Input=3D Point Cloud2024.05 | 53 | 52.55 | 52.78 | |
| PointLLM-13BInput=3D Point Cloud, Zero-shot=true2023.08 | 53 | 52.55 | 53.39 | |
| Point-Bind LLMLLM Size=7B, Input=3D Point Cloud2024.05 | 51.9 | 39.71 | 45.81 | |
| Point-Bind LLMInput=3D Point Cloud, Zero-shot=true2023.08 | 51.9 | 39.71 | 25.53 | |
| PointLLM-13BLLM Size=13B, 3D Data Size=730K, Input=Point Cloud2026.02 | 51.7 | 52.67 | 52.19 | |
| PointLLM-7BLLM Size=7B, 3D Data Size=730K, Input=Point Cloud2026.02 | 51.34 | 50.36 | 50.85 | |
| Point-Bind LLMLLM Size=7B, 3D Data Size=-, Input=Point Cloud2026.02 | 46.6 | 45.02 | 45.81 | |
| LLAVA-7BLLM Size=7B, Trainable Params=7.03B, Input=Single-V. Img.2024.05 | 39.75 | 39.67 | 39.71 | |
| LLaVA-7BInput=Single-V. Img., Zero-shot=true2023.08 | 39.75 | 39.67 | 44.86 | |
| LLAVA-13BLLM Size=13B, Trainable Params=13.03B, Input=Single-V. Img.2024.05 | 37.12 | 36.06 | 36.59 | |
| LLaVA-13BInput=Single-V. Img., Zero-shot=true2023.08 | 37.12 | 36.06 | 44.17 | |
| LLaVA-13BLLM Size=13B, 3D Data Size=0K, Input=Single-V. Img.2026.02 | 27.71 | 27.76 | 27.74 | |
| LLaVA-7BLLM Size=7B, 3D Data Size=0K, Input=Single-V. Img.2026.02 | 27.11 | 21.68 | 24.4 | |
| InstructBLIP-13BLLM Size=13B, Trainable Params=0.20B, Input=Single-V. Img.2024.05 | 25.97 | 31.4 | 28.69 | |
| InstructBLIP-13BInput=Single-V. Img., Zero-shot=true2023.08 | 25.97 | 31.4 | 31.47 | |
| ShapeLLM-13Bzero-shot=true, parameters=13B, protocol=PointLLM2026.05 | 22.17 | 20.15 | — | |
| GPT-4o miniLLM Size=-, 3D Data Size=0K, Input=Single-V. Img.2026.02 | 22 | 23.1 | 22.55 | |
| InstructBLIP-13BLLM Size=13B, 3D Data Size=0K, Input=Single-V. Img.2026.02 | 21.56 | 21.92 | 21.74 | |
| GPT4PointLLM Size=2.7B, 3D Data Size=660K, Input=Point Cloud2026.02 | 21.39 | 21.07 | 21.23 | |
| ShapeLLM-7Bzero-shot=true, parameters=7B, protocol=PointLLM2026.05 | 19.8 | 11.84 | — | |
| InstructBLIP-7BLLM Size=7B, Trainable Params=0.20B, Input=Single-V. Img.2024.05 | 19.53 | 31.48 | 25.51 | |
| InstructBLIP-7BInput=Single-V. Img., Zero-shot=true2023.08 | 19.53 | 31.48 | 34.5 | |
| InstructBLIP-7BLLM Size=7B, 3D Data Size=0K, Input=Single-V. Img.2026.02 | 17.67 | 22.81 | 20.24 | |
| ShapeLLM-13BLLM Size=13B, Trainable Params=13.04B, Input=3D Point Cloud2024.05 | — | — | 52.96 | |
| ShapeLLM-7BLLM Size=7B, Trainable Params=7.04B, Input=3D Point Cloud2024.05 | — | — | 53.08 |