arXiv cs.AI· Ioannis N. Ziogas, Aamna Al Shehhi, Ahsan H. Khandoker, Leontios J. Hadjileontiadis·· 11 小时前AI 评分24
变分分解自编码（VDA）提升潜在表征的解耦性
Variational decomposition autoencoding improves disentanglement of latent representations
AI 导读
研究者提出变分分解自编码（VDA）框架，通过变分分解自编码器（DecVAE）在分布层面进行潜在自编码，学习与时频特征对齐的多个潜在子空间。在模拟数据及语音识别、构音障碍语音评估、情感语音分类三个公开数据集上，DecVAE 在解耦质量和鲁棒性上优于基于 VAE 的方法，且当生成因子具有明显频谱特征时潜在子空间呈现因子特异性。
正文 · 原文
Abstract:Understanding the structure of complex, nonstationary, high-dimensional time-evolving signals is a central challenge in scientific data analysis. In many domains, such as speech and biomedical signal processing, the ability to learn disentangled and interpretable representations is critical for uncovering latent generative mechanisms. Traditional approaches to unsupervised representation learning, including variational autoencoders (VAEs), often struggle to capture the temporal and spectral diversity inherent in such data. Here we introduce variational decomposition autoencoding (VDA), a framework that incorporates a strong structural bias toward signal decomposition into variational representation learning. VDA is instantiated through variational decomposition autoencoders (DecVAEs), i.e., encoder-only neural networks that perform latent autoencoding at the level of distributions to learn multiple latent subspaces aligned with time-frequency characteristics. We evaluate DecVAEs on simulated data and three publicly available scientific datasets, spanning speech recognition, dysarthric speech evaluation, and emotional speech classification. DecVAEs improve on VAE-based methods in disentanglement quality and robustness, and their latent subspaces become factor-specific when the generative factors have distinct spectral signatures. These findings position decomposition-aware architectures as potential tools for extracting structured representations from speech, and potentially from other dynamic signals with factors exhibiting distinct spectral signatures, with applications in clinical diagnostics, human-computer interaction, and adaptive neurotechnologies.
| Comments: | Supplementary information file at: this https URL |
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Audio and Speech Processing (eess.AS); Signal Processing (eess.SP); Machine Learning (stat.ML) |
| Cite as: | arXiv:2601.06844 [cs.LG] |
| (or arXiv:2601.06844v2 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2601.06844 arXiv-issued DOI via DataCite |
Submission history
From: Ioannis Ziogas [view email]
[v1]
Sun, 11 Jan 2026 10:16:34 UTC (10,222 KB)
[v2]
Fri, 30 Oct 2026 16:15:53 UTC (26,345 KB)
来源：arXiv cs.AI · arxiv.org