Speech Content Extraction on AudioMarathon 1.0 (test)
12.4SERBaichuan-Omni-1.5
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Baichuan-Omni-1.52026.02 | 12.4 | 11.2 | 86.5 | |
| Aero-1-Audio2026.02 | 17.9 | 56.6 | 43.7 | |
| Phi-4-Multimodal2026.02 | 18.4 | 69.3 | 92.7 | |
| Gemma-3n-E4B-itModel Size=4B2026.02 | 19 | 56.9 | 93.2 | |
| Audio-Flamingo-32026.02 | 21.7 | 78.9 | 94.3 | |
| Gemma-3n-E2B-itModel Size=2B2026.02 | 22.5 | 51.6 | 91.3 | |
| Voxtral-Mini-3B-2507Parameters=3B, Version=25072026.02 | 24.3 | 71.1 | 96.8 | |
| Qwen2.5-Omni-3BParameters=3B2026.02 | 25.2 | 82.3 | 94.7 | |
| GPT-4o-Audio (2024-12-17)Release Date=2024-12-172026.02 | 25.7 | 60.2 | 94.7 | |
| GPT-4o-Audio (2024-10-01)Release Date=2024-10-012026.02 | 25.8 | 61.4 | 94.4 | |
| Qwen2.5-Omni-7BParameters=7B2026.02 | 26.3 | 85.1 | 98.1 | |
| Audio-Flamingo-22026.02 | 26.8 | 39.8 | 1 | |
| Gemini-2.5-Flash2026.02 | 28.1 | 83.6 | 96.8 | |
| Gemini-2.5-Flash-Lite2026.02 | 30.3 | 64 | 96.5 | |
| Speech-XL2026.02 | 31.1 | 75.6 | 96.1 |