Video Question Answering On How2Qa
评估指标
Accuracy
评测结果
各个模型在此基准测试上的表现结果
模型名称 | Accuracy | Paper Title | Repository |
|---|---|---|---|
| Text + Text (no Multimodal Pretext Training) | 93.2 | Towards Fast Adaptation of Pretrained Contrastive Models for Multi-channel Video-Language Retrieval | |
| Just Ask | 84.4 | Just Ask: Learning to Answer Questions from Millions of Narrated Videos | |
| FrozenBiLM | 86.7 | Zero-Shot Video Question Answering via Frozen Bidirectional Language Models | |
| ATP | 65.1 | Revisiting the "Video" in Video-Language Understanding | |
| SeViLA | 83.7 | - | - |
| Just Ask (0-shot) | 51.1 | Just Ask: Learning to Answer Questions from Millions of Narrated Videos | |
| Hero w/ pre-training | 77.75 | HERO: Hierarchical Encoder for Video+Language Omni-representation Pre-training | |
| FrozenBiLM (0-shot) | 58.4 | Zero-Shot Video Question Answering via Frozen Bidirectional Language Models |
0 of 8 row(s) selected.