We report an overall RMIS score across all datasets, where the higher the score is, the better the model is!
We warmly welcome fellow researchers to test their models on the RMIS benchmark! If you would like to showcase your model on the leaderboard, please provide a minimum inference code and contact us at jab22@mails.tsinghua.edu.cn
| Model | Variant | Size | RMIS | Anomaly Detection | Fault Diagnosis | |||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| DCASE | Mean | IICA | IIEE | WTPG | MaFaulDa | SDUST | UMGED | PU | Mean | |||||||||||||||
| 20 | 21 | 22 | 23 | 24 | 25 | sound | vib | bearing | gear | sound | vib | vol | cur | vib | cur | |||||||||
| Wav2Vec 2.0 | base | 95M | 53.27 | 64.76 | 55.14 | 55.77 | 55.69 | 52.76 | 55.69 | 56.64 | 46.77 | 96.61 | 60.81 | 60.72 | 78.34 | 54.12 | 95.01 | 9.08 | 9.34 | 11.93 | 15.58 | 67.00 | 43.42 | 49.90 |
| Wav2Vec 2.0 | large | 315M | 50.72 | 65.61 | 55.23 | 54.44 | 48.09 | 50.51 | 33.98 | 51.31 | 45.49 | 72.48 | 78.33 | 62.83 | 87.35 | 49.98 | 97.04 | 10.12 | 8.40 | 9.18 | 12.50 | 69.24 | 48.86 | 50.14 |
| Wav2Vec 2.0 | 1B | 962M | 52.46 | 65.50 | 55.22 | 55.81 | 54.12 | 51.83 | 51.60 | 55.68 | 44.23 | 94.97 | 84.84 | 49.40 | 77.08 | 49.95 | 96.72 | 8.39 | 10.89 | 8.49 | 9.93 | 64.59 | 40.69 | 49.24 |
| Wav2Vec 2.0 | 2B | 2.16B | 53.71 | 65.49 | 55.47 | 56.35 | 55.47 | 53.91 | 54.53 | 56.87 | 42.09 | 78.81 | 84.49 | 59.96 | 84.14 | 50.31 | 98.24 | 8.37 | 12.06 | 10.80 | 11.13 | 72.67 | 44.07 | 50.55 |
| Whisper | tiny | 7.6M | 50.03 | 62.85 | 54.42 | 53.72 | 54.09 | 52.53 | 54.01 | 55.27 | 40.36 | 55.16 | 71.86 | 44.40 | 76.92 | 51.17 | 96.99 | 6.71 | 9.52 | 9.48 | 12.51 | 66.00 | 41.12 | 44.78 |
| Whisper | base | 19.8M | 52.36 | 65.66 | 54.89 | 56.12 | 56.90 | 53.70 | 56.14 | 57.24 | 40.10 | 77.24 | 83.95 | 36.35 | 80.82 | 51.30 | 96.90 | 8.98 | 9.71 | 9.02 | 10.21 | 69.95 | 42.70 | 47.48 |
| Whisper | small | 87M | 51.40 | 64.02 | 54.44 | 54.85 | 55.50 | 53.79 | 54.67 | 56.21 | 41.12 | 70.21 | 83.55 | 38.72 | 76.75 | 49.23 | 95.84 | 8.63 | 10.50 | 8.68 | 9.89 | 68.89 | 43.57 | 46.58 |
| Whisper | medium | 306M | 51.63 | 65.41 | 54.76 | 55.58 | 56.02 | 54.13 | 55.21 | 56.85 | 41.28 | 66.84 | 85.62 | 37.31 | 76.66 | 48.93 | 96.11 | 9.29 | 11.79 | 8.09 | 9.61 | 67.65 | 44.02 | 46.40 |
| Whisper | large-v3 | 635M | 51.78 | 62.27 | 53.65 | 54.51 | 53.97 | 52.53 | 54.54 | 55.25 | 47.57 | 70.44 | 87.24 | 59.32 | 73.36 | 50.85 | 94.08 | 9.14 | 10.10 | 8.34 | 8.05 | 68.80 | 40.68 | 48.31 |
| PaSST | base | 86M | 57.69 | 66.60 | 56.84 | 57.58 | 59.56 | 57.64 | 57.76 | 59.33 | 60.50 | 97.56 | 92.69 | 59.30 | 88.59 | 61.86 | 97.11 | 7.16 | 9.80 | 10.64 | 15.87 | 74.92 | 52.62 | 56.05 |
| AudioMAE | base | 85M | 51.61 | 63.87 | 54.86 | 56.34 | 57.96 | 55.30 | 56.82 | 57.52 | 58.88 | 53.89 | 64.02 | 47.48 | 80.95 | 59.74 | 98.18 | 6.91 | 10.93 | 10.01 | 11.59 | 56.02 | 35.54 | 45.70 |
| BEATs | iter3 | 90M | 57.30 | 74.20 | 60.50 | 58.95 | 63.31 | 56.93 | 59.09 | 62.16 | 77.93 | 68.17 | 70.57 | 52.87 | 84.45 | 62.37 | 98.23 | 5.98 | 10.32 | 10.95 | 17.23 | 74.18 | 48.42 | 52.44 |
| OpenBEATs | base | 90M | 55.67 | 70.86 | 58.60 | 57.45 | 62.48 | 56.10 | 57.32 | 60.47 | 77.87 | 70.93 | 63.60 | 45.50 | 83.39 | 57.31 | 96.51 | 5.47 | 9.51 | 11.09 | 13.78 | 76.96 | 49.50 | 50.88 |
| OpenBEATs | large | 310M | 56.95 | 72.78 | 59.93 | 58.29 | 63.29 | 57.32 | 58.03 | 61.61 | 75.18 | 82.86 | 62.33 | 42.69 | 79.37 | 70.65 | 96.86 | 6.91 | 9.66 | 12.63 | 12.48 | 78.57 | 49.67 | 52.30 |
| EAT | base30 | 88M | 57.62 | 72.69 | 58.03 | 58.84 | 59.35 | 56.48 | 58.65 | 60.67 | 64.48 | 96.65 | 61.28 | 68.32 | 91.90 | 64.03 | 95.46 | 7.89 | 9.39 | 12.49 | 18.87 | 68.33 | 50.38 | 54.57 |
| EAT | large10 | 310M | 58.03 | 73.81 | 57.06 | 57.94 | 60.29 | 57.78 | 60.34 | 61.20 | 65.80 | 94.06 | 68.91 | 58.47 | 88.42 | 62.64 | 96.32 | 9.57 | 11.60 | 11.22 | 15.54 | 75.74 | 54.73 | 54.85 |
| CED | tiny | 5.5M | 55.36 | 66.99 | 56.17 | 56.75 | 60.09 | 56.40 | 58.40 | 59.15 | 47.93 | 74.23 | 89.43 | 53.77 | 84.56 | 59.02 | 96.06 | 8.09 | 12.96 | 10.31 | 12.65 | 71.50 | 49.95 | 51.57 |
| CED | mini | 9.6M | 55.95 | 67.48 | 56.35 | 56.59 | 60.05 | 57.44 | 58.26 | 59.36 | 50.21 | 84.26 | 89.26 | 56.95 | 84.88 | 59.24 | 95.50 | 8.36 | 11.66 | 10.13 | 12.43 | 71.50 | 48.56 | 52.53 |
| CED | small | 22M | 56.01 | 67.50 | 56.65 | 56.87 | 60.76 | 57.88 | 58.15 | 59.63 | 50.49 | 80.23 | 91.20 | 54.67 | 84.67 | 58.00 | 96.12 | 8.05 | 12.15 | 10.26 | 13.14 | 72.89 | 49.08 | 52.38 |
| CED | base | 86M | 56.71 | 67.60 | 56.67 | 56.95 | 60.99 | 57.89 | 58.45 | 59.76 | 52.26 | 90.08 | 89.20 | 57.64 | 85.95 | 60.20 | 95.84 | 8.12 | 11.57 | 10.03 | 13.63 | 72.14 | 50.91 | 53.66 |
| DaSheng | base | 85M | 56.37 | 69.25 | 57.28 | 56.02 | 60.95 | 56.39 | 55.82 | 59.28 | 74.42 | 99.10 | 46.84 | 63.53 | 92.06 | 61.37 | 98.30 | 5.64 | 4.45 | 11.74 | 20.08 | 74.93 | 42.42 | 53.45 |
| DaSheng | 0.6B | 630M | 55.43 | 68.35 | 56.76 | 55.35 | 59.85 | 56.34 | 55.41 | 58.68 | 75.14 | 99.25 | 45.41 | 57.26 | 91.11 | 58.23 | 97.59 | 5.27 | 4.28 | 11.69 | 14.68 | 75.78 | 42.63 | 52.18 |
| DaSheng | 1.2B | 1130M | 55.91 | 69.52 | 57.06 | 55.82 | 61.14 | 56.32 | 56.01 | 59.31 | 73.18 | 99.00 | 48.89 | 58.16 | 90.79 | 55.50 | 97.99 | 5.42 | 4.52 | 11.76 | 17.32 | 75.99 | 44.11 | 52.51 |
| MERT | v1-95M | 95M | 52.96 | 65.54 | 55.40 | 54.68 | 54.46 | 53.37 | 52.74 | 56.03 | 58.59 | 98.07 | 54.93 | 42.47 | 77.55 | 57.92 | 98.87 | 9.26 | 4.42 | 11.55 | 16.38 | 75.47 | 43.01 | 49.88 |
| MERT | v1-330M | 315M | 56.70 | 68.79 | 56.87 | 55.50 | 58.34 | 55.17 | 55.65 | 58.39 | 64.11 | 99.85 | 69.02 | 66.88 | 92.38 | 61.71 | 99.19 | 5.54 | 3.92 | 15.02 | 18.57 | 72.97 | 46.12 | 55.02 |
| MuQ | msd | 333M | 53.65 | 66.91 | 56.75 | 56.04 | 57.69 | 55.64 | 56.09 | 58.19 | 62.23 | 95.33 | 37.98 | 41.22 | 88.08 | 60.67 | 96.38 | 6.11 | 5.72 | 15.07 | 9.00 | 78.57 | 42.20 | 49.12 |
| Qwen2-Audio | - | 637M | 54.73 | 67.32 | 55.38 | 56.46 | 57.19 | 56.07 | 56.27 | 58.12 | 45.59 | 73.56 | 89.91 | 59.63 | 83.35 | 60.13 | 97.12 | 7.63 | 10.26 | 9.17 | 10.63 | 73.17 | 47.46 | 51.35 |
| Qwen2.5-Omni | - | 640M | 55.40 | 70.94 | 57.39 | 57.20 | 59.14 | 56.83 | 57.05 | 59.76 | 45.49 | 66.56 | 90.73 | 53.52 | 86.21 | 60.49 | 97.60 | 9.02 | 11.35 | 9.25 | 11.87 | 74.84 | 46.74 | 51.05 |
| Audio Flamingo 3 | - | 637M | 55.12 | 67.02 | 55.06 | 57.08 | 57.52 | 56.76 | 56.55 | 58.33 | 47.41 | 88.57 | 92.41 | 40.85 | 87.20 | 58.27 | 98.09 | 7.69 | 11.56 | 9.30 | 10.21 | 73.37 | 49.77 | 51.90 |
| MiDaShengLM | - | 630M | 53.65 | 67.10 | 55.46 | 56.10 | 58.70 | 56.15 | 56.03 | 58.26 | 63.78 | 59.63 | 86.39 | 35.69 | 76.94 | 68.31 | 98.25 | 7.75 | 15.54 | 9.20 | 7.29 | 67.47 | 41.28 | 49.04 |
| Time-MoE | base | 113M | 49.03 | 55.43 | 51.83 | 53.72 | 51.38 | 50.35 | 50.58 | 52.22 | 56.84 | 77.21 | 76.31 | 47.83 | 73.75 | 53.92 | 92.52 | 5.29 | 3.81 | 12.71 | 6.52 | 56.67 | 32.52 | 45.84 |
| Sundial | base | 116M | 50.32 | 63.87 | 54.50 | 53.15 | 49.06 | 51.84 | 52.17 | 54.10 | 50.53 | 98.41 | 55.57 | 15.56 | 59.94 | 45.01 | 86.57 | 5.22 | 4.54 | 37.30 | 26.10 | 68.73 | 51.52 | 46.54 |
| PEEMD | - | - | 46.16 | 58.84 | 53.25 | 53.32 | 53.88 | 53.63 | 50.63 | 53.93 | 35.40 | 66.41 | 61.29 | 51.23 | 54.30 | 30.45 | 53.14 | 10.85 | 10.55 | 14.58 | 10.69 | 63.64 | 36.49 | 38.39 |
| TFPred | wc1 | 3.9M | 41.88 | 58.41 | 53.22 | 51.39 | 51.08 | 50.66 | 51.14 | 52.65 | 35.35 | 77.15 | 24.91 | 10.94 | 36.30 | 26.29 | 56.56 | 9.71 | 9.53 | 22.48 | 10.80 | 51.52 | 33.02 | 31.12 |
| LiConvFormer | - | 320K | 43.99 | 61.65 | 53.35 | 51.96 | 49.50 | 51.72 | 50.62 | 53.13 | 33.13 | 75.10 | 34.08 | 13.56 | 47.15 | 34.00 | 71.24 | 7.52 | 6.78 | 25.68 | 11.92 | 56.60 | 36.23 | 34.85 |
| BearLLM | - | 204K | 45.75 | 64.12 | 53.12 | 48.95 | 44.04 | 51.96 | 50.47 | 52.11 | 26.29 | 66.25 | 65.76 | 45.54 | 48.99 | 44.11 | 77.84 | 12.66 | 9.66 | 11.33 | 5.59 | 65.34 | 32.83 | 39.40 |
| RotLLM | - | 770K | 47.70 | 62.91 | 53.55 | 53.80 | 53.63 | 51.16 | 50.94 | 54.33 | 56.98 | 85.82 | 48.33 | 15.65 | 38.65 | 60.10 | 92.24 | 6.89 | 8.13 | 9.36 | 5.00 | 69.79 | 37.02 | 41.07 |
| CoWS | - | 6.31M | 51.52 | 59.98 | 53.93 | 53.12 | 52.39 | 52.07 | 50.03 | 53.59 | 44.72 | 74.40 | 86.45 | 80.43 | 76.17 | 48.40 | 92.96 | 10.68 | 4.79 | 21.04 | 9.29 | 59.79 | 33.80 | 49.46 |
| ECHO | tiny | 5.5M | 60.67 | 70.43 | 59.01 | 58.86 | 63.92 | 55.77 | 56.55 | 60.76 | 90.31 | 100.00 | 65.81 | 58.84 | 92.26 | 76.70 | 98.78 | 12.90 | 21.34 | 19.92 | 25.77 | 74.21 | 50.87 | 60.59 |
| ECHO | small | 22M | 60.31 | 72.50 | 60.20 | 59.13 | 64.25 | 56.76 | 57.45 | 61.71 | 90.34 | 99.95 | 60.45 | 61.81 | 93.18 | 69.08 | 98.54 | 14.17 | 16.96 | 14.90 | 18.34 | 75.70 | 52.37 | 58.91 |
| FISHER | tiny | 5.5M | 61.12 | 70.86 | 58.76 | 56.40 | 58.62 | 53.64 | 56.37 | 59.11 | 85.31 | 98.72 | 83.84 | 75.79 | 92.58 | 71.81 | 99.20 | 11.90 | 23.58 | 21.10 | 28.24 | 72.36 | 56.21 | 63.13 |
| FISHER | mini | 10M | 61.43 | 70.19 | 58.40 | 57.62 | 61.07 | 54.59 | 55.75 | 59.60 | 86.02 | 99.05 | 84.74 | 75.74 | 93.20 | 70.09 | 98.55 | 13.69 | 20.52 | 22.95 | 29.97 | 72.01 | 55.73 | 63.25 |
| FISHER | small | 22M | 62.23 | 71.04 | 59.48 | 59.64 | 62.63 | 55.62 | 58.46 | 61.15 | 90.23 | 98.96 | 86.77 | 72.61 | 95.86 | 76.35 | 99.08 | 12.72 | 18.35 | 17.42 | 25.52 | 74.90 | 54.29 | 63.31 |
■ Speech Encoders
■ Audio Encoders
■ Music Encoders
■ LALM Audio Encoders
■ Time Series Encoders
■ Signal Encoders
■ Our Model (FISHER)