Video-to-Audio temporal grounding on ChronusAV
3.9BLEU-4PadCaptioner
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| PadCaptionerModel Parameters=3B2026.07 | 3.9 | 5.3 | 3.8 | 15 | |
| ChronusOmniModel Size=7B2025.12 | 3.61 | 4.9 | 3.27 | 13.6 | |
| ChronusOmniModel Parameters=7B2026.07 | 3.6 | 4.9 | 3.3 | 13.6 | |
| Qwen2.5-OmniModel Size=7B2025.12 | 0.54 | 0.91 | 0.61 | 0.3 | |
| Qwen2.5-OmniModel Parameters=7B2026.07 | 0.5 | 0.9 | 0.6 | 0.3 | |
| Qwen3-OmniModel Parameters=30B-A3B2026.07 | 0.4 | 1.6 | 1.3 | 0.7 | |
| Qwen3-OmniModel Size=30B-A3B2025.12 | 0.35 | 1.58 | 1.25 | 0.74 | |
| OlaModel Parameters=7B2026.07 | 0.2 | 1.2 | 0.5 | 0.5 | |
| OlaModel Size=7B2025.12 | 0.15 | 1.15 | 0.53 | 0.45 | |
| ARC-Hunyuan-VideoModel Size=7B2025.12 | 0.12 | 0.94 | 0.7 | 0.1 | |
| LongVALE-LLMModel Size=7B2025.12 | 0.1 | 1.26 | 0.64 | 0.24 | |
| VideoLLaMAModel Parameters=7B2026.07 | 0.1 | 1.1 | 0.8 | 0.1 | |
| LongVALE-LLMModel Parameters=7B2026.07 | 0.1 | 1.3 | 0.6 | 0.2 | |
| ARC-HY-VideoModel Parameters=7B2026.07 | 0.1 | 0.9 | 0.7 | 0.1 | |
| VideoLLaMAModel Size=7B2025.12 | 0.08 | 1.11 | 0.81 | 0.06 | |
| AVicunaModel Size=7B2025.12 | 0.02 | 0.54 | 0.23 | 0.1 | |
| AVicunaModel Parameters=7B2026.07 | 0 | 0.5 | 0.2 | 0.1 |