الملخص
Pretrained audio encoders are reused for downstream tasks that are often unknown when the encoder is trained, so their usefulness depends partly on which signal properties survive the pretext objective. We study this retained information through paired source reconstruction. Using a shared Stable Audio Open latent diffusion decoder, we reconstruct five-second, 44.1-kHz stereo music from frozen representations produced by supervised classifiers (VGGish, ConvNeXt), an audio-text contrastive model (CLAP), and a waveform-reconstruction model (EnCodec). These objectives impose different pressures to preserve source detail, while their exposed interfaces vary substantially in temporal and spectral resolution. Evaluating on the Million Song Dataset (MSD), we find clear differences in reconstructability across encoder families, while within encoder comparisons show improved recovery when finer temporal or spectral structure is exposed. Even compressed task oriented embeddings support reconstructions that preserve measurable source specificity and high level musical content.
الكلمات المفتاحية
الموضوع
بيانات النشر
- المجلة
- غير متاح
- وصول مفتوح
- وصول مفتوح أخضر
اقتبس هذه المقالة
APA 7
Glytsos, M., & McFee, B. (2026). How Much Audio Is Left In An Embedding? An Inversion Audit Of Audio Encoders. https://omanscience.com/ar/articles/how-much-audio-is-left-in-an-embedding-an-inversion-audit-of-audio-encoders
MLA 9
Glytsos, Marios, and Brian McFee. "How Much Audio Is Left In An Embedding? An Inversion Audit Of Audio Encoders." https://omanscience.com/ar/articles/how-much-audio-is-left-in-an-embedding-an-inversion-audit-of-audio-encoders.
شيكاغو (المؤلف–التاريخ)
Glytsos, Marios, and Brian McFee. 2026. "How Much Audio Is Left In An Embedding? An Inversion Audit Of Audio Encoders." https://omanscience.com/ar/articles/how-much-audio-is-left-in-an-embedding-an-inversion-audit-of-audio-encoders.
هارفارد
Glytsos, M. and McFee, B. (2026) 'How Much Audio Is Left In An Embedding? An Inversion Audit Of Audio Encoders', Available at: https://omanscience.com/ar/articles/how-much-audio-is-left-in-an-embedding-an-inversion-audit-of-audio-encoders.
فانكوفر
Glytsos M, McFee B. How Much Audio Is Left In An Embedding? An Inversion Audit Of Audio Encoders. https://omanscience.com/ar/articles/how-much-audio-is-left-in-an-embedding-an-inversion-audit-of-audio-encoders
IEEE
M. Glytsos, and B. McFee, "How Much Audio Is Left In An Embedding? An Inversion Audit Of Audio Encoders," https://omanscience.com/ar/articles/how-much-audio-is-left-in-an-embedding-an-inversion-audit-of-audio-encoders.