arXiv:2611.10192 · v1
Lightweight Neural Prediction for MDCT-Based Perceptual Audio Coding
摘要
Perceptual audio codecs often use predictive coding to further reduce signal redundancy. Rather than learning the entire codec pipeline end-to-end, we insert a small neural network as a drop-in, deterministic predictor in an otherwise conventional codec: the network estimates the next Modified Discrete Cosine Transform (MDCT) coefficient block from a short context of reconstructed blocks, and the encoder transmits the quantized residual. We compare Multilayer Perceptron (MLP) and Convolutional Neural Network (CNN) predictors against a non-predictive baseline and a per-line Normalized Least Mean Squares (NLMS) linear predictor (similar to the MPEG-2 AAC Main Profile predictor). The CNN consistently outperforms the MLP, the linear predictor, and the baseline in signal-to-noise ratio (SNR), noise-to-mask ratio (NMR), and MUSHRA evaluation scores. Using a 512-sample analysis window (w = 512), our neural predictors recover, and the CNN exceeds, the coded quality obtained by the non-predictive baseline after quadrupling the analysis-window length to w = 2048, at substantially lower algorithmic delay. The CNN uses only 0.13M parameters (roughly 20x fewer than the MLP) and has a decoder real-time factor (RTF) of ~0.10 on a consumer CPU.
作者备注:10 pages, 2 figures, 9 tables. Presented at the 161st AES Convention, Nashville, TN, USA, October 30-November 1, 2026 (published version: https://aes.org/publications/elibrary/elibrary-page/?id=23480). Audio samples and supplemental materials: https://github.com/NiccoloAbate/neural-predictive-audio-coding
期刊信息:Proc. AES Show 2026 Nashville (161st AES Convention), Oct. 2026, Paper 50