Music Genre Classification on GTZAN
95.5AccuracyAudio Flamingo 3 + CoAT
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Audio Flamingo 3 + CoATBackbone=Audio Flamingo 3, Method=CoAT2026.06 | 95.5 | — | |
| Audio Flamingo 3Backbone=Audio Flamingo 3, Method=Baseline2026.06 | 94.99 | — | |
| UniWhisperEvaluation Protocol=MLP2026.02 | 94.5 | — | |
| Qwen2.5-Omni-7B + CoATBackbone=Qwen2.5-Omni-7B, Method=CoAT2026.06 | 92.99 | — | |
| Qwen2.5-Omni-7BBackbone=Qwen2.5-Omni-7B, Method=Baseline2026.06 | 92.89 | — | |
| Qwen2-Audio + CoATBackbone=Qwen2-Audio, Method=CoAT2026.06 | 92.39 | — | |
| Qwen2-AudioBackbone=Qwen2-Audio, Method=Baseline2026.06 | 90.29 | — | |
| Ensemble (Submission)Ensemble Strategy=Embedding Concatenation2026.01 | 89.8 | — | |
| BEATsEvaluation Protocol=MLP2026.02 | 89.8 | — | |
| UniWhisperEvaluation Protocol=kNN2026.02 | 89.5 | — | |
| Dasheng 1.2BParameter Count=1.2B2026.01 | 88.6 | — | |
| MATPAC++Role=Teacher, Size=86M2026.04 | 87.6 | — | |
| MobileNetV3Teacher=None (Supervised), Size=2.9M2026.04 | 87.4 | — | |
| Challenge BaselineDescription=Best among Dasheng-base, data2vec, and Whisper2026.01 | 86.9 | — | |
| BEATs (300M) BalancedParameter Count=300M, Pre-training Mixture=Balanced (40:30:30)2026.01 | 85.9 | — | |
| S-SONDO (DyMN)Teacher=MATPAC++, Size=8.7M2026.04 | 85.6 | — | |
| M2DNumber of Parameters=86M, Masking ratio=0.7, Target Feeding=all patches, Linear Evaluation=true2024.04 | 85.2 | — | |
| BEATsEvaluation Protocol=kNN2026.02 | 85.2 | — | |
| S-SONDO (MobileNetV3)Teacher=MATPAC++, Size=2.9M2026.04 | 85.2 | — | |
| HTS-ATNumber of Parameters=31M, Linear Evaluation=true2024.04 | 85.1 | — | |
| BEATSiter3+Number of Parameters=90M, Linear Evaluation=true2024.04 | 84.6 | — | |
| BEATs (300M) SpeechParameter Count=300M, Pre-training Mixture=Speech-heavy (70:15:15)2026.01 | 84.5 | — | |
| ASTNumber of Parameters=86M, Linear Evaluation=true2024.04 | 84.3 | — | |
| M2DRole=Teacher, Size=86M2026.04 | 84.3 | — | |
| CLAP-UEvaluation Protocol=MLP2026.02 | 84.2 | — | |
| M2DNumber of Parameters=86M, Masking ratio=0.7, Target Feeding=masked patches, Linear Evaluation=true2024.04 | 84.1 | — | |
| BEATs (90M) iter3Parameter Count=90M, Iteration=32026.01 | 83.6 | — | |
| M2DNumber of Parameters=86M, Masking ratio=0.6, Target Feeding=masked patches, Linear Evaluation=true2024.04 | 83.4 | — | |
| Audio-Flamingo 3Size=8.3B2026.04 | 83.2 | — | |
| AST-Fusion#5#12Number of Parameters=86M, Linear Evaluation=true2024.04 | 82.9 | — | |
| Whisper-U-3Evaluation Protocol=MLP2026.02 | 82.3 | — | |
| S-SONDO (MobileNetV3)Teacher=M2D, Size=2.9M2026.04 | 81.4 | — | |
| M2DNumber of Parameters=86M, Masking ratio=0.6, Target Feeding=all patches, Linear Evaluation=true2024.04 | 81.3 | — | |
| DyMNTeacher=None (Supervised), Size=8.7M2026.04 | 80.8 | — | |
| CLAP-UEvaluation Protocol=kNN2026.02 | 79.5 | — | |
| MSM-MAENumber of Parameters=86M, Masking ratio=0.75, Linear Evaluation=true2024.04 | 79.2 | — | |
| S-SONDO (DyMN)Teacher=M2D, Size=8.7M2026.04 | 79.2 | — | |
| ATST BaseNumber of Parameters=86M, Linear Evaluation=true2024.04 | 78.9 | — | |
| S-SONDO (ERes2Net)Teacher=MATPAC++, Size=1.4M2026.04 | 77.6 | — | |
| S-SONDO (ERes2Net)Teacher=M2D, Size=1.4M2026.04 | 77.6 | — | |
| Qwen2-Audio-InstructSize=8.4B2026.04 | 77.2 | — | |
| DeLoRes-MNumber of Parameters=5.3M, Linear Evaluation=true2024.04 | 75 | — | |
| WavLMEvaluation Protocol=MLP2026.02 | 74 | — | |
| MiDashengLMSize=8.3B2026.04 | 72.7 | — | |
| BEATSiter3Number of Parameters=90M, Linear Evaluation=true2024.04 | 72.6 | — | |
| WhisperEvaluation Protocol=MLP2026.02 | 71.2 | — | |
| Whisper-UEvaluation Protocol=MLP2026.02 | 70.5 | — | |
| BYOL-ANumber of Parameters=5.3M, Linear Evaluation=true2024.04 | 70.1 | — | |
| HuBERTEvaluation Protocol=MLP2026.02 | 69.6 | — | |
| Audio-Flamingo 2Size=4.4B2026.04 | 69.1 | — | |
| TinyMUSize=229M(2.7%)2026.04 | 65.7 | — | |
| Whisper-U-3Evaluation Protocol=kNN2026.02 | 64.1 | — | |
| wav2vec 2.0Evaluation Protocol=MLP2026.02 | 63.7 | — | |
| ERes2NetTeacher=None (Supervised), Size=1.4M2026.04 | 62.8 | — | |
| LisGANEvaluation Protocol=Zero-shot2024.12 | 58.24 | 0.0287 | |
| ZeroDiffusionEvaluation Protocol=Zero-shot2024.12 | 58.03 | 0.0281 | |
| MusiLingoSize=7.1B2026.04 | 57.7 | — | |
| ALEEvaluation Protocol=Zero-shot2024.12 | 57.05 | 0.0109 | |
| CADA-VAEEvaluation Protocol=Zero-shot2024.12 | 56.5 | 0.0463 | |
| WhisperEvaluation Protocol=kNN2026.02 | 53.3 | — | |
| WavLMEvaluation Protocol=kNN2026.02 | 43.6 | — | |
| CLAPEvaluation Protocol=MLP2026.02 | 40.4 | — | |
| MU-LLaMASize=7.7B2026.04 | 37.3 | — | |
| CLAPEvaluation Protocol=kNN2026.02 | 34.9 | — | |
| wav2vec 2.0Evaluation Protocol=kNN2026.02 | 31.1 | — | |
| HuBERTEvaluation Protocol=kNN2026.02 | 20.7 | — | |
| MellowSize=167M2026.04 | 16.5 | — | |
| Whisper-UEvaluation Protocol=kNN2026.02 | 11.2 | — |