PointLLM-7B V1.2

Reported on 6 benchmarks across 1 task · 1 paper · 1 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Computer Vision6 results

3D Object CaptioningonObjaverse
Precision· 2023-08-31
82.14
best: 83.14 (MiniGPT-3D)
SOTA
PointLLM: Empowering Large Language Models to Understand Point Clouds arXiv:2308.16911
3D Object CaptioningonObjaverse
Sentence-BERT· 2023-08-31
47.47
best: 49.54 (MiniGPT-3D)
PointLLM: Empowering Large Language Models to Understand Point Clouds arXiv:2308.16911
3D Object CaptioningonObjaverse
Correctness· 2023-08-31
3.04
best: 3.5 (MiniGPT-3D)
PointLLM: Empowering Large Language Models to Understand Point Clouds arXiv:2308.16911
3D Object CaptioningonObjaverse
GPT-4· 2023-08-31
44.85
best: 57.06 (MiniGPT-3D)
PointLLM: Empowering Large Language Models to Understand Point Clouds arXiv:2308.16911
3D Object CaptioningonObjaverse
Hallucination· 2023-08-31
0.66
best: 1.16 (3D-LLM)
PointLLM: Empowering Large Language Models to Understand Point Clouds arXiv:2308.16911
3D Object CaptioningonObjaverse
SimCSE· 2023-08-31
48.55
best: 51.39 (MiniGPT-3D)
PointLLM: Empowering Large Language Models to Understand Point Clouds arXiv:2308.16911