RECORD
Acoustic Word Embeddings for Untranscribed Target Languages with Continued Pretraining and Learned Pooling
Computation and Language · Submitted 3 Jun 2023
abstract compiled · 2026-09-23 · no language-model review
The authors' main claim Evaluating on four languages, we show that both methods outperform a recent approach on word discrimination.
The compiled record
The paper's own sentences, arranged. Each label is drawn from a closed vocabulary; the number beside it is the probability the classifier gave that label. Numbers in the tables are the authors', checked by code. No sentence in this record was written by a model.
Open the compiled record
The abstract, sentence by sentence
Acoustic word embeddings are typically created by training a pooling function using pairs of word-like units. For unsupervised systems, these are mined using k-nearest neighbor (KNN) search, which is slow. Recently, mean-pooled representations from a pre-trained self-supervised English model were suggested as a promising alternative, but their performance on target languages was not fully competitive. Here, we explore improvements to both approaches: we use continued pre-training to adapt the self-supervised model to the target language, and we use a multilingual phone recognizer (MPR) to mine phone n-gram pairs for training the pooling function. Evaluating on four languages, we show that both methods outperform a recent approach on word discrimination. Moreover, the MPR method is orders of magnitude faster than KNN, and is highly data efficient. We also show a small improvement from performing learned pooling on top of the continued pre-trained representations.
background contribution method result interpretation main claim
| # | sentence | role | the authors' own work | states the main conclusion | strength |
|---|---|---|---|---|---|
| s0 | Acoustic word embeddings are typically created by training a pooling function using pairs of word-like units. | background 1.00 | 0.05 | 0.02 | asserted 0.99 |
| s1 | For unsupervised systems, these are mined using k-nearest neighbor (KNN) search, which is slow. | background 0.92 | 0.06 | 0.02 | asserted 1.00 |
| s2 | Recently, mean-pooled representations from a pre-trained self-supervised English model were suggested as a promising alternative, but their performance on target languages was not fully competitive. | background 0.53 | 0.05 | 0.03 | hedged possibility 0.52 |
| s3 | Here, we explore improvements to both approaches: we use continued pre-training to adapt the self-supervised model to the target language, and we use a multilingual phone recognizer (MPR) to mine phone n-gram pairs for training the pooling function. | contribution 0.56 | 0.89 | 0.09 | asserted 0.83 |
| s4 | Evaluating on four languages, we show that both methods outperform a recent approach on word discrimination. | result 1.00 | 0.96 | 0.54 | asserted 1.00 |
| s5 | Moreover, the MPR method is orders of magnitude faster than KNN, and is highly data efficient. | result 1.00 | 0.94 | 0.18 | asserted 1.00 |
| s6 | We also show a small improvement from performing learned pooling on top of the continued pre-trained representations. | result 1.00 | 0.96 | 0.09 | asserted 1.00 |
What was examined, and what was not
- arXiv listing abstract v1
- full text
- figures and tables
- references and citing sentences
- relations to other papers
- P abstract
- S sections
- T body sentences
- C citations
- R cross-paper relations
The paper
Original abstract
Acoustic word embeddings are typically created by training a pooling function using pairs of word-like units. For unsupervised systems, these are mined using k-nearest neighbor (KNN) search, which is slow. Recently, mean-pooled representations from a pre-trained self-supervised English model were suggested as a promising alternative, but their performance on target languages was not fully competitive. Here, we explore improvements to both approaches: we use continued pre-training to adapt the self-supervised model to the target language, and we use a multilingual phone recognizer (MPR) to mine phone n-gram pairs for training the pooling function. Evaluating on four languages, we show that both methods outperform a recent approach on word discrimination. Moreover, the MPR method is orders of magnitude faster than KNN, and is highly data efficient. We also show a small improvement from performing learned pooling on top of the continued pre-trained representations.
Human review
How the literature uses it
1 arXiv paper cite this in 1 sentence
Discussion
Know something about this paper? Leave a comment, or highlight any passage above to comment on it.