# Acoustic Feature Learning via Deep Variational Canonical Correlation Analysis

arXiv: https://arxiv.org/abs/1708.04673 · HTML: https://mlsanity.com/paper/1708.04673/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 11 Aug 2017.

## Built on by (most cited first, 4 of 4)

- [Identifiability Results for Multimodal Contrastive Learning](https://mlsanity.com/paper/2303.09166/) (2023)
- [Multi-view autoencoders for Fake News Detection](https://mlsanity.com/paper/2504.08102/) (2025)
- [Variational Interpretable Learning from Multi-view Data](https://mlsanity.com/paper/2202.13503/) (2022)
- [Speech representation learning: Learning bidirectional encoders with single-view, multi-view, and multi-task methods](https://mlsanity.com/paper/2308.00129/) (2023)

## Abstract (arXiv)

We study the problem of acoustic feature learning in the setting where we have access to another (non-acoustic) modality for feature learning but not at test time. We use deep variational canonical correlation analysis (VCCA), a recently proposed deep generative method for multi-view representation learning. We also extend VCCA with improved latent variable priors and with adversarial learning. Compared to other techniques for multi-view feature learning, VCCA's advantages include an intuitive latent variable interpretation and a variational lower bound objective that can be trained end-to-end efficiently. We compare VCCA and its extensions with previous feature learning methods on the University of Wisconsin X-ray Microbeam Database, and show that VCCA-based feature learning improves over previous methods for speaker-independent phonetic recognition.