Automated Audio Captioning on Clotho
55.92AAC ScoreKimi-Audio
Evaluation Results
| Method | Links | |
|---|---|---|
| Kimi-AudioSize=7B2025.09 | 55.92 | |
| SLAM-OmniSize=0.5B2025.09 | 54.52 | |
| GLM-4-VoiceSize=9B2025.09 | 13.15 | |
| TtTSize=3B2025.09 | 12.63 | |
| Pretrain+TtTSize=3B2025.09 | 11.55 | |
| Qwen2.5-3B (AR)Size=3B2025.09 | 9.73 | |
| Qwen2.5-3B (NAR)Size=3B2025.09 | 9.54 | |
| VITA-AudioSize=7B2025.09 | 6.18 | |
| MoshiSize=7B2025.09 | 4.32 | |
| Mini-OmniSize=0.5B2025.09 | 3.61 | |
| LLaMA-OmniSize=8B2025.09 | 2.53 | |
| SpeechGPTSize=7B2025.09 | 2.1 |