ATP

Reported on 2 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.

Reasoning2 results

Video Question AnsweringonNExT-QA
Accuracy· 2022-06-03
54.3
best: 85.5 (LinVT-Qwen2-VL (7B))
SOTA
Revisiting the "Video" in Video-Language Understanding arXiv:2206.01720
Video Question AnsweringonHow2QA
Accuracy· 2022-06-03
65.1
best: 93.2 (Text + Text (no Multimodal Pretext Training))
Revisiting the "Video" in Video-Language Understanding arXiv:2206.01720