VidCtx (7B)

Reported on 6 benchmarks across 2 tasks · 1 paper

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

Natural Language Processing3 results

Question AnsweringonNExT-QA
Accuracy· 2024-12-23
70.7
best: 79.6 (VideoMultiAgent (GPT-4o))
VidCtx: Context-aware Video Question Answering with Image Models arXiv:2412.17415
Question AnsweringonSTAR Benchmark
Accuracy· 2024-12-23
51.1
best: 59 (VideoChat2)
VidCtx: Context-aware Video Question Answering with Image Models arXiv:2412.17415
Question AnsweringonIntentQA
Accuracy· 2024-12-23
67.1
best: 71.5 (ENTER)
VidCtx: Context-aware Video Question Answering with Image Models arXiv:2412.17415

Reasoning3 results

Video Question AnsweringonNExT-QA
Accuracy· 2024-12-23
70.7
best: 85.5 (LinVT-Qwen2-VL (7B))
VidCtx: Context-aware Video Question Answering with Image Models arXiv:2412.17415
Video Question AnsweringonSTAR Benchmark
Accuracy· 2024-12-23
51.1
best: 59 (VideoChat2)
VidCtx: Context-aware Video Question Answering with Image Models arXiv:2412.17415
Video Question AnsweringonIntentQA
Accuracy· 2024-12-23
67.1
best: 71.5 (ENTER)
VidCtx: Context-aware Video Question Answering with Image Models arXiv:2412.17415