The Equalizer: Introducing Shape-Gain Decomposition in Neural Audio Codecs

2026-02-17

AI Summary

This work proposes a novel approach that systematically integrates the classical shape–gain decomposition principle from speech coding into neural audio codec frameworks. Existing neural audio codecs exhibit sensitivity to input amplitude variations when jointly encoding signal gain and shape, leading to codebook redundancy and degraded rate–distortion performance. To address this, the proposed method first decomposes the signal into short-term gain and normalized shape components, which are then encoded separately using scalar quantization and a neural network, respectively, before being jointly reconstructed. This strategy substantially enhances robustness to amplitude fluctuations, achieves superior rate–distortion performance on speech signals, significantly reduces model complexity, and seamlessly integrates into diverse neural audio codec architectures.

Technology Category

• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Machine Learning: Deep Generative Models & Autoencoders
• Natural Language Processing: Speech

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Economics, Online Markets and Human Computation: Incentives in network design for Web infrastructures and ecosystems
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation

Abstract

Neural audio codecs (NACs) typically encode the short-term energy (gain) and normalized structure (shape) of speech/audio signals jointly within the same latent space. As a result, they are poorly robust to a global variation of the input signal level in the sense that such variation has strong influence on the embedding vectors at the output of the encoder and their quantization. This methodology is inherently inefficient, leading to codebook redundancy and suboptimal bitrate-distortion performance. To address these limitations, we propose to introduce shape-gain decomposition, widely used in classical speech/audio coding, into the NAC framework. The principle of the proposed Equalizer methodology is to decompose the input signal -- before the NAC encoder -- into gain and normalized shape vector on a short-term basis. The shape vector is processed by the NAC, while the gain is quantized with scalar quantization and transmitted separately. The output (decoded) signal is reconstructed from the normalized output of the NAC and the quantized gain. Our experiments conducted on speech signals show that this general methodology, easily applicable to any NAC, enables a substantial gain in bitrate-distortion performance, as well as a massive reduction in complexity.

Problem

• neural audio codecs
• shape-gain decomposition
• bitrate-distortion performance
• codebook redundancy
• signal level robustness

Innovation

• shape-gain decomposition
• neural audio codecs
• bitrate-distortion optimization
• scalar quantization
• Equalizer