Audio-to-Video temporal grounding on ChronusAV
1.4BLEU-4PadCaptioner
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| PadCaptionerModel Parameters=3B2026.07 | 1.4 | 3.7 | 2.6 | 3.7 | |
| ChronusOmniModel Size=7B2025.12 | 1.01 | 3.17 | 2.12 | 3.03 | |
| ChronusOmniModel Parameters=7B2026.07 | 1 | 3.2 | 2.1 | 3 | |
| Qwen3-OmniModel Size=30B-A3B2025.12 | 0.22 | 1.72 | 1.49 | 0.39 | |
| LongVALE-LLMModel Size=7B2025.12 | 0.21 | 1.88 | 1.02 | 0.87 | |
| OlaModel Parameters=7B2026.07 | 0.2 | 1.6 | 1.1 | 0.4 | |
| LongVALE-LLMModel Parameters=7B2026.07 | 0.2 | 1.9 | 1 | 0.9 | |
| Qwen2.5-OmniModel Parameters=7B2026.07 | 0.2 | 1.6 | 1 | 0.6 | |
| Qwen3-OmniModel Parameters=30B-A3B2026.07 | 0.2 | 1.7 | 1.5 | 0.4 | |
| Qwen2.5-OmniModel Size=7B2025.12 | 0.19 | 1.61 | 0.97 | 0.58 | |
| OlaModel Size=7B2025.12 | 0.18 | 1.6 | 1.08 | 0.4 | |
| AVicunaModel Size=7B2025.12 | 0.11 | 1.45 | 0.65 | 0.38 | |
| ARC-Hunyuan-VideoModel Size=7B2025.12 | 0.11 | 1.12 | 0.91 | 0.11 | |
| VideoLLaMAModel Parameters=7B2026.07 | 0.1 | 1.2 | 0.9 | 0 | |
| AVicunaModel Parameters=7B2026.07 | 0.1 | 1.5 | 0.7 | 0.4 | |
| ARC-HY-VideoModel Parameters=7B2026.07 | 0.1 | 1.1 | 0.9 | 0.1 | |
| VideoLLaMAModel Size=7B2025.12 | 0.07 | 1.22 | 0.93 | 0 |