List your company
ResearchFormerly WizWand

Audio Classification on AudioSet-2M (full)

48.6mAPBEATSiter3+
-1.4510411.5429824.53737.53102Jul 13, 2022Oct 11, 2022Jan 10, 2023Apr 10, 2023Jul 10, 2023Oct 8, 2023Jan 7, 2024
Updated 8mo ago

Evaluation results

MethodLinks
BEATSiter3+#Param=90M, Pre-training Data=AS, Pre-training Category=Self-Supervised, Knowledge Distillation=extra pseudo-labels2024.0148.6
EAT#Param=88M, Pre-training Data=AS, Pre-training Category=Ours2024.0148.6
BEATSiter2#Param=90M, Pre-training Data=AS, Pre-training Category=Self-Supervised2024.0148.1
BEATSiter3#Param=90M, Pre-training Data=AS, Pre-training Category=Self-Supervised2024.0148
BEATSiter1#Param=90M, Pre-training Data=AS, Pre-training Category=Self-Supervised2024.0147.9
Audio-MAE (local)Backbone=ViT-B, PT-Data=AS2022.0747.3
Audio-MAE#Param=86M, Pre-training Data=AS, Pre-training Category=Self-Supervised2024.0147.3
MaskSpecBackbone=ViT-B, PT-Data=AS2022.0747.1
HTS-ATBackbone=Swin-B, PT-Data=IN, Additional Supervised Training=AS-2M2022.0747.1
PaSSTBackbone=DeiT-B, PT-Data=IN, Additional Supervised Training=AS-2M2022.0747.1
PassT#Param=86M, Pre-training Data=IN, Pre-training Category=Supervised2024.0147.1
HTS-AT#Param=31M, Pre-training Data=IN, Pre-training Category=Supervised2024.0147.1
MaskSpec#Param=86M, Pre-training Data=AS, Pre-training Category=Self-Supervised2024.0147.1
Audio-MAE (global)Backbone=ViT-B, PT-Data=AS2022.0746.8
ASTBackbone=DeiT-B, PT-Data=IN2022.0745.9
AST#Param=86M, Pre-training Data=IN, Pre-training Category=Supervised2024.0145.9
ERANNBackbone=CNN, PT-Data=None2022.0745
PSLABackbone=EffNet, PT-Data=IN2022.0744.4
PSLA#Param=14M, Pre-training Data=IN, Pre-training Category=Supervised2024.0144.4
MBTBackbone=ViT-B, PT-Data=IN-21K2022.0744.3
MBT#Param=86M, Pre-training Data=IN-21K, Pre-training Category=Supervised2024.0144.3
PANNBackbone=CNN, PT-Data=None2022.0743.1
PANN#Param=81M, Pre-training Data=Not Specified, Pre-training Category=Supervised2024.0143.1
ConformerBackbone=Conformer, PT-Data=AS2022.0741.1
Conformer#Param=88M, Pre-training Data=AS, Pre-training Category=Self-Supervised2024.0141.1
AudioCLIP#Param=93M, Pre-training Data=TI+AS, Pre-training Category=Supervised2024.0125.9
BEATSSL Data=AS, ensemble_size=10 models2022.120.506
BEATSSL Data=AS, ensemble_size=5 models2022.120.504
PaSSTSL Data=IN+AS2022.120.496
HTS-ATSL Data=IN+AS2022.120.487
ASTSL Data=IN+AS2022.120.485
PSLASL Data=IN+AS2022.120.474