Establishing degrees of closeness between audio recordings along different dimensions using large-scale cross-lingual models
This work addresses the challenge of analyzing speech representations in low-resource language studies, offering a method for comparative work on under-documented languages, though it is incremental in nature.
The authors tackled the problem of understanding what information multilingual speech model representations encode by proposing an unsupervised ABX test method to measure closeness between audio recordings across dimensions like room acoustics, linguistic genre, and phonetics. They found that longer audio vectors better discriminate extra-linguistic characteristics, while shorter snippets are more effective for segmental information.
In the highly constrained context of low-resource language studies, we explore vector representations of speech from a pretrained model to determine their level of abstraction with regard to the audio signal. We propose a new unsupervised method using ABX tests on audio recordings with carefully curated metadata to shed light on the type of information present in the representations. ABX tests determine whether the representations computed by a multilingual speech model encode a given characteristic. Three experiments are devised: one on room acoustics aspects, one on linguistic genre, and one on phonetic aspects. The results confirm that the representations extracted from recordings with different linguistic/extra-linguistic characteristics differ along the same lines. Embedding more audio signal in one vector better discriminates extra-linguistic characteristics, whereas shorter snippets are better to distinguish segmental information. The method is fully unsupervised, potentially opening new research avenues for comparative work on under-documented languages.