English

Learning Joint Acoustic-Phonetic Word Embeddings

Machine Learning 2019-08-02 v1 Computation and Language Sound Audio and Speech Processing Machine Learning

Abstract

Most speech recognition tasks pertain to mapping words across two modalities: acoustic and orthographic. In this work, we suggest learning encoders that map variable-length, acoustic or phonetic, sequences that represent words into fixed-dimensional vectors in a shared latent space; such that the distance between two word vectors represents how closely the two words sound. Instead of directly learning the distances between word vectors, we employ weak supervision and model a binary classification task to predict whether two inputs, one of each modality, represent the same word given a distance threshold. We explore various deep-learning models, bimodal contrastive losses, and techniques for mining hard negative examples such as the semi-supervised technique of self-labeling. Our best model achieves an F1F_1 score of 0.95 for the binary classification task.

Keywords

Cite

@article{arxiv.1908.00493,
  title  = {Learning Joint Acoustic-Phonetic Word Embeddings},
  author = {Mohamed El-Geish},
  journal= {arXiv preprint arXiv:1908.00493},
  year   = {2019}
}

Comments

8 pages, 4 figures

R2 v1 2026-06-23T10:37:29.950Z