List your company

Full-audio Detection on SONICS full audio

99.99RecallFusion Segment Transformer
74.707681.271387.83594.3987Jan 20, 2026

Evaluation results

MethodLinks
Fusion Segment TransformerFeature Extractor=MERT2026.0199.9999.9910099.9999.99100
MERT (Standalone)Architecture=Baseline (Linear Head)2026.0199.9699.9610099.9610099.96
Segment TransformerFeature Extractor=MERT2026.0199.9699.9299.8899.9299.9999.87
Fusion Segment TransformerFeature Extractor=Music2vec2026.0199.8899.9499.9999.9310099.99
Fusion Segment TransformerFeature Extractor=FXencoder2026.0199.6499.7499.8199.7399.9899.83
Segment TransformerFeature Extractor=FXencoder2026.0199.6199.5599.5399.5799.9599.48
Fusion Segment TransformerFeature Extractor=Muffin2026.0197.7598.2898.6398.1999.7898.76
SpecTTTra-αArchitecture=Baseline2026.0196——97—99
Segment TransformerFeature Extractor=Music2vec2026.0195.4695.5596.0995.7799.0795.66
ConvNeXtArchitecture=Baseline2026.0195——96—98
EfficientViTArchitecture=Baseline2026.0194——95—97
Segment TransformerFeature Extractor=Wav2vec 2.02026.0187.2674.7371.2778.4674.1260.75
SpecTTTra-βArchitecture=Baseline2026.0186——92—99
ViTArchitecture=Baseline2026.0182——89—98
SpecTTTra-γArchitecture=Baseline2026.0179——88—99
Fusion Segment TransformerFeature Extractor=Wav2vec 2.02026.0175.6877.5877.0476.3585.6379.33