# Multi-band Frequency Reconstruction for Neural Psychoacoustic Coding

arXiv: https://arxiv.org/abs/2505.07235 · HTML: https://mlsanity.com/paper/2505.07235/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 12 May 2025.

## Built on by (most cited first, 2 of 2)

- [Towards Interactive Intelligence for Digital Humans](https://mlsanity.com/paper/2512.13674/) (2025)
- [Fusion Segment Transformer: Bi-Directional Attention Guided Fusion Network for AI-Generated Music Detection](https://mlsanity.com/paper/2601.13647/) (2026)

## Abstract (arXiv)

Achieving high-fidelity audio compression while preserving perceptual quality across diverse content remains a key challenge in Neural Audio Coding (NAC). We introduce MUFFIN, a fully convolutional Neural Psychoacoustic Coding (NPC) framework that leverages psychoacoustically guided multi-band frequency reconstruction. At its core is a Multi-Band Spectral Residual Vector Quantization (MBS-RVQ) module that allocates bitrate across frequency bands based on perceptual salience. This design enables efficient compression while disentangling speaker identity from content using distinct codebooks. MUFFIN incorporates a transformer-inspired convolutional backbone and a modified snake activation to enhance resolution in fine-grained spectral regions. Experimental results on multiple benchmarks demonstrate that MUFFIN consistently outperforms existing approaches in reconstruction quality. A high-compression variant achieves a state-of-the-art 12.5 Hz rate with minimal loss. MUFFIN also proves effective in downstream generative tasks, highlighting its promise as a token representation for integration with language models. Audio samples and code are available.