HILCodec: High-Fidelity and Lightweight Neural Audio Codec

TL;DR

HILCodec is a novel neural audio codec that achieves high-fidelity, low-bitrate audio compression with a lightweight model and real-time streaming capability, addressing previous model complexity and distortion issues.

Contribution

The paper introduces HILCodec, a neural audio codec with a variance-constrained design and a distortion-free discriminator, improving quality and efficiency over prior models.

Findings

HILCodec outperforms existing codecs in audio quality across various bitrates.

The variance-constrained design stabilizes performance as network depth increases.

The distortion-free discriminator reduces waveform distortions in reconstructed audio.

Abstract

The recent advancement of end-to-end neural audio codecs enables compressing audio at very low bitrates while reconstructing the output audio with high fidelity. Nonetheless, such improvements often come at the cost of increased model complexity. In this paper, we identify and address the problems of existing neural audio codecs. We show that the performance of the SEANet-based codec does not increase consistently as the network depth increases. We analyze the root cause of such a phenomenon and suggest a variance-constrained design. Also, we reveal various distortions in previous waveform domain discriminators and propose a novel distortion-free discriminator. The resulting model, HILCodec, is a real-time streaming audio codec that demonstrates state-of-the-art quality across various bitrates and audio types.