Zero-shot Audio Classification Explanation on ESC50 contamination
39.4AILMAC-ZS
Evaluation results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| LMAC-ZSTraining Set=Full, Masking Domain=STFT-Masking, Classification Accuracy=58.6%2024.05 | 39.4 | 8.28 | 11.81 | 0.69 | 0.8 | 0.67 | 11.79 | 0.09 | |
| LMAC-ZSTraining Set=CT, Masking Domain=STFT-Masking, Classification Accuracy=58.6%2024.05 | 35.65 | 12.23 | 13.04 | 0.69 | 0.74 | 0.53 | 12.18 | 0.09 | |
| GradCam++Masking Domain=STFT-Masking, Classification Accuracy=64%2024.05 | 34.13 | 12.24 | 10.84 | 0.67 | 0.72 | 0.41 | 12.44 | 0.34 | |
| GradCam++Masking Domain=STFT-Masking, Classification Accuracy=58.6%2024.05 | 29.52 | 14.84 | 10.17 | 0.7 | 0.7 | 0.39 | 12.48 | 0.35 | |
| LMAC-ZSTraining Set=CT, Masking Domain=STFT-Masking, Classification Accuracy=64%2024.05 | 25.77 | 17.79 | 9.67 | 0.63 | 0.63 | 0.61 | 11.96 | 0.04 | |
| LMAC-ZSTraining Set=Full, Masking Domain=STFT-Masking, Classification Accuracy=64%2024.05 | 25.73 | 15.9 | 7.23 | 0.66 | 0.62 | 0.72 | 11.47 | 0.05 | |
| GradCamMasking Domain=STFT-Masking, Classification Accuracy=64%2024.05 | 24.93 | 22.91 | 12.78 | 0.67 | 0.5 | 0.7 | 11.72 | 0.18 | |
| LMAC-ZSTraining Set=CT, Masking Domain=Mel-Masking, Classification Accuracy=64.8%2024.05 | 24.38 | 20.69 | 11.29 | 0.43 | 0.56 | 0.8 | 9.26 | 0.11 | |
| GradCamMasking Domain=STFT-Masking, Classification Accuracy=58.6%2024.05 | 23.77 | 25.25 | 12.24 | 0.69 | 0.49 | 0.69 | 11.73 | 0.17 | |
| LMAC-ZSTraining Set=Full, Masking Domain=Mel-Masking, Classification Accuracy=57.2%2024.05 | 20.43 | 21.57 | 9.71 | 0.42 | 0.54 | 0.82 | 9.08 | 0.15 | |
| LMAC-ZSTraining Set=CT, Masking Domain=Mel-Masking, Classification Accuracy=57.2%2024.05 | 19.25 | 24.3 | 8.83 | 0.4 | 0.49 | 0.81 | 9.18 | 0.13 | |
| IGMasking Domain=STFT-Masking, Classification Accuracy=58.6%2024.05 | 16.37 | 25.67 | 7.21 | 0.7 | 0.51 | 0.71 | 11.73 | 0.011 | |
| IGMasking Domain=STFT-Masking, Classification Accuracy=64%2024.05 | 15.55 | 27.15 | 6.51 | 0.66 | 0.46 | 0.73 | 11.67 | 0.01 | |
| GradCam++Masking Domain=Mel-Masking, Classification Accuracy=64.8%2024.05 | 12.85 | 32.81 | 6.5 | 0.47 | 0.32 | 0.41 | 10.36 | 0.35 | |
| SmoothGradMasking Domain=STFT-Masking, Classification Accuracy=58.6%2024.05 | 11.8 | 30.63 | 5.15 | 0.7 | 0.42 | 0.52 | 12.06 | 0.0002 | |
| GradCam++Masking Domain=Mel-Masking, Classification Accuracy=57.2%2024.05 | 9.82 | 35.81 | 4.53 | 0.42 | 0.29 | 0.39 | 10.4 | 0.35 | |
| SmoothGradMasking Domain=STFT-Masking, Classification Accuracy=64%2024.05 | 9.18 | 29.6 | 3.91 | 0.67 | 0.4 | 0.53 | 12.05 | 0 | |
| LMAC-ZSTraining Set=Full, Masking Domain=Mel-Masking, Classification Accuracy=64.8%2024.05 | 8.95 | 30.55 | 3.69 | 0.38 | 0.35 | 0.86 | 8.79 | 0.1 | |
| GradCamMasking Domain=Mel-Masking, Classification Accuracy=57.2%2024.05 | 6.78 | 40.71 | 3.13 | 0.29 | 0.19 | 0.69 | 9.66 | 0.18 | |
| GradCamMasking Domain=Mel-Masking, Classification Accuracy=64.8%2024.05 | 6.5 | 39.05 | 3.06 | 0.33 | 0.2 | 0.7 | 9.66 | 0.18 | |
| SmoothGradMasking Domain=Mel-Masking, Classification Accuracy=64.8%2024.05 | 0.63 | 47.4 | 0.17 | 0.03 | 0.02 | 0.28 | 10.55 | 0.04 | |
| SmoothGradMasking Domain=Mel-Masking, Classification Accuracy=57.2%2024.05 | 0.62 | 48.55 | 0.13 | 0.024 | 0.022 | 0.29 | 10.54 | 0.039 | |
| IGMasking Domain=Mel-Masking, Classification Accuracy=57.2%2024.05 | 0.55 | 48.88 | 0.091 | 0.073 | 0.02 | 0.56 | 10.13 | 0.039 | |
| IGMasking Domain=Mel-Masking, Classification Accuracy=64.8%2024.05 | 0.53 | 47.7 | 0.1 | 0.1 | 0.01 | 0.56 | 10.12 | 0.04 |