3D-LLM (BLIP2-flant5)

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.

Natural Language Processing6 results

Visual Question Answering (VQA)onScanQA Test w/ objects
BLEU-1· 2023-07-24
38.3
best: 39.73 (NaviLLM)
SOTA
3D-LLM: Injecting the 3D World into Large Language Models arXiv:2307.12981
Visual Question Answering (VQA)onScanQA Test w/ objects
BLEU-4· 2023-07-24
11.6
best: 24.06 (BridgeQA)
3D-LLM: Injecting the 3D World into Large Language Models arXiv:2307.12981
Visual Question Answering (VQA)onScanQA Test w/ objects
CIDEr· 2023-07-24
69.6
best: 103.1 (LLaVA-3D)
3D-LLM: Injecting the 3D World into Large Language Models arXiv:2307.12981
Visual Question Answering (VQA)onScanQA Test w/ objects
Exact Match· 2023-07-24
19.1
best: 31.29 (BridgeQA)
3D-LLM: Injecting the 3D World into Large Language Models arXiv:2307.12981
Visual Question Answering (VQA)onScanQA Test w/ objects
METEOR· 2023-07-24
14.9
best: 20.8 (LLaVA-3D)
3D-LLM: Injecting the 3D World into Large Language Models arXiv:2307.12981
Visual Question Answering (VQA)onScanQA Test w/ objects
ROUGE· 2023-07-24
35.3
best: 49.6 (LLaVA-3D)
3D-LLM: Injecting the 3D World into Large Language Models arXiv:2307.12981