List your company
ResearchFormerly WizWand

Audio Classification on US8K (test)

0.9807R@1 AccuracyEAT (LoRA) + Joint Training
0.0647720.3025610.540350.778139Jun 7, 2023Dec 7, 2023Jun 7, 2024Dec 8, 2024Jun 9, 2025Dec 9, 2025Jun 11, 2026
Updated 4mo ago

Evaluation results

MethodLinks
EAT (LoRA) + Joint TrainingTraining Protocol=Joint Training, PEFT Strategy=LoRA2026.020.9807
PACE2026.020.9749
RanPAC (High-Order)2026.020.9708
RanPAC (High-Order) w/o FSAFeature Space Alignment=false2026.020.9649
ACL (High-Order)2026.020.9598
BEATs (iter3+)Pre-train Data=AS, Training Protocol=Fine-tuning2026.060.911
PaSST (IN+AS)Pre-train Data=IN+AS, Training Protocol=Fine-tuning2026.060.901
AST (Single)Pre-train Data=IN+AS, Training Protocol=Fine-tuning2026.060.898
MaskSpecPre-train Data=AS, Training Protocol=Fine-tuning2026.060.896
MAE-ASTPre-train Data=AS+LS, Training Protocol=Fine-tuning2026.060.884
AudioPGPre-train Data=Proc. Audio, Training Protocol=Fine-tuning2026.060.8817
MSM-MAEPre-train Data=AS, Training Protocol=Fine-tuning2026.060.881
M2DProtocol=Linear evaluation2023.060.876
PANNs (CNN14)Pre-train Data=AudioSet (Sup.), Training Protocol=Fine-tuning2026.060.874
SSASTPre-train Data=AS+LS, Training Protocol=Fine-tuning2026.060.865
ATST-ClipProtocol=Linear evaluation2023.060.858
ATST-FrameProtocol=Linear evaluation2023.060.858
BYOL-APre-train Data=AudioSet, Training Protocol=Fine-tuning2026.060.848
HiDe-Prompt2026.020.7989
BYOL-A-V2Protocol=Linear evaluation2023.060.797
BYOL-AProtocol=Linear evaluation2023.060.791
Ishikawa et al. (Shaders1k)Pre-train Data=Syn. Images, Training Protocol=Fine-tuning2026.060.783
HiDe-Adapter2026.020.7803
ViT-B (ImageNet SL)Pre-train Data=IN-1k, Training Protocol=Fine-tuning2026.060.776
HiDe-LoRA2026.020.7648
Supervised (Scratch)Pre-train Data=-, Training Protocol=Fine-tuning2026.060.7534
Audio Flamingozero-shot=true2024.020.75
SKMPre-train Data=-, Training Protocol=Fine-tuning2026.060.74
CLAPPre-training Dataset=AudioCaps, Zero-shot=true2023.100.72
Deshmukh et al., 2023zero-shot=true2024.020.719
CyCLAPPre-training Dataset=Clotho, Zero-shot=true2023.100.7182
CyCLAPPre-training Dataset=Clotho, Zero-shot=true2023.100.7046
CLAPsPre-training Dataset=Clotho, Zero-shot=true2023.100.7027
CLAPPre-training Dataset=AudioCaps, Zero-shot=true2023.100.6978
CLAPPre-training Dataset=Clotho, Zero-shot=true2023.100.6964
CyCLAPPre-training Dataset=AudioCaps, Zero-shot=true2023.100.6946
CyCLAPsPre-training Dataset=AudioCaps, Zero-shot=true2023.100.6938
CyCLAPsPre-training Dataset=Clotho, Zero-shot=true2023.100.6875
CyCLAPPre-training Dataset=AudioCaps, Zero-shot=true2023.100.679
CLAPPre-training Dataset=Clotho, Zero-shot=true2023.100.6776
CLAPSPre-training Dataset=Clotho, Zero-shot=true2023.100.6742
CyCLAPSPre-training Dataset=AudioCaps, Zero-shot=true2023.100.6694
CLAPsPre-training Dataset=AudioCaps, Zero-shot=true2023.100.6649
CyCLAPSPre-training Dataset=Clotho, Zero-shot=true2023.100.6598
CLAPSPre-training Dataset=AudioCaps, Zero-shot=true2023.100.6414
LoRASub2026.020.5781
EAT (Frozen) + Linear probeEvaluation Protocol=Linear probe2026.020.4999
EAT (LoRA) + Linear ProbePEFT Strategy=LoRA, Evaluation Protocol=Linear Probe2026.020.4968
EAT (Prompt) + Linear ProbePEFT Strategy=Prompt, Evaluation Protocol=Linear Probe2026.020.4498
S-Prompt++2026.020.4257
Nearest Class Mean (NCM) w/ FSAFeature Space Alignment=true2026.020.4244
DualPrompt2026.020.424
EAT (Adapter) + Linear ProbePEFT Strategy=Adapter, Evaluation Protocol=Linear Probe2026.020.3876
L2P2026.020.3875
Nearest Class Mean (NCM)2026.020.3609
Random Guess (1/K)Pre-train Data=-, Training Protocol=Fine-tuning2026.060.1