Video Captioning / Summarization on Vatex
28.8Rouge-Lv-LCM
Evaluation results
| Method | Links | ||
|---|---|---|---|
| v-LCMvariant=v-LCM2026.03 | 28.8 | 48.7 | |
| Qwen2-VLModel Size=7B2026.03 | 28.5 | 51.6 | |
| Qwen2.5-VLModel Size=7B2026.03 | 27.5 | 50.8 | |
| LCMvariant=LCM2026.03 | 23.8 | 30.8 | |
| InterVL2Model Size=1B2026.03 | 22.4 | 42.4 | |
| InternVL-2.5Model Size=8B2026.03 | 20.6 | 42.4 | |
| Percep. LMModel Size=3B2026.03 | 19.3 | 30.8 | |
| InterVL2Model Size=4B2026.03 | 19.2 | 42.4 | |
| InternVL-2.5Model Size=1B2026.03 | 19.2 | 40.3 | |
| Percep. LMModel Size=1B2026.03 | 19.1 | 29.6 | |
| Percep. LMModel Size=8B2026.03 | 19 | 30.8 | |
| InternVL-2.5Model Size=4B2026.03 | 17.3 | 36.2 | |
| Qwen2-VLModel Size=2B2026.03 | 16.4 | 30.8 | |
| InterVL2Model Size=8B2026.03 | 16.2 | 42.4 | |
| Qwen2.5-VLModel Size=3B2026.03 | 15 | 27.6 |