Multimodal Summarization on MM-AVS
6.02BLEU-4CoE
Evaluation Results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| CoEzero-shot=true, training-free=true, backbone=Qwen2.5-VL-7B-Instruct2026.03 | 6.02 | 19.53 | 17.89 | 14.51 | 88.62 | 33.91 | 4.01 | |
| TCoTzero-shot=true, training-free=true, backbone=Qwen2.5-VL-7B-Instruct2026.03 | 5.03 | 15.98 | 7.14 | 10.17 | 85.69 | 24.47 | 3.31 | |
| CoFzero-shot=true, training-free=true, backbone=InternVL3-8B2026.03 | 4.98 | 14.17 | 7.49 | 9.89 | 82.99 | 19.49 | 2.47 | |
| ViTCoTzero-shot=true, training-free=true, backbone=Qwen2.5-VL-7B-Instruct2026.03 | 4.26 | 15.88 | 4.32 | 13.1 | 84.28 | 26.8 | 3.24 | |
| CoSzero-shot=true, training-free=true, backbone=LLaVA-Video / LongVA2026.03 | 4.1 | 16.18 | 4.45 | 14.26 | 83.93 | 26.25 | 2.89 |