AI RESEARCH
Reconstruct! Don't Encode: Self-Supervised Representation Reconstruction Loss for High-Intelligibility and Low-Latency Streaming Neural Audio Codec
arXiv CS.AI
•
ArXi:2603.05887v1 Announce Type: cross Neural audio codecs optimized for mel-spectrogram reconstruction often fail to preserve intelligibility. While semantic encoder distillation improves encoded representations, it does not guarantee content preservation in reconstructed speech. In this work, we nstrate that self-supervised representation reconstruction (SSRR) loss fundamentally improves codec