LLaVA-Video

Reported on 4 benchmarks across 3 tasks · 1 paper · 1 SOTA

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

Reasoning2 results

Video Question AnsweringonNExT-QA
Accuracy· 2024-10-03
83.2
best: 85.5 (LinVT-Qwen2-VL (7B))
SOTA
Video Instruction Tuning With Synthetic Data arXiv:2410.02713
Video Question AnsweringonZero-shot Video Question Answering on LongVideoBench
Accuracy (% )· uses extra data· 2024-10-03
61.9
best: 66.7 (Gemini 1.5 Pro)
Video Instruction Tuning With Synthetic Data arXiv:2410.02713

Natural Language Processing2 results

Question AnsweringonZero-shot Video Question Answering on LongVideoBench
Accuracy (% )· uses extra data· 2024-10-03
61.9
best: 66.7 (Gemini 1.5 Pro)
Video Instruction Tuning With Synthetic Data arXiv:2410.02713
Visual Question Answering (VQA)onSQA3D
Exact Match· 2024-10-03
48.5
best: 60.1 (LLaVA-3D)
Video Instruction Tuning With Synthetic Data arXiv:2410.02713