Behavioral Analysis of Pathological Speaker Embeddings of Patients During Oncological Treatment of Oral Cancer

TL;DR

This study investigates how speaker embeddings of oral cancer patients change during treatment, revealing significant voice characteristic shifts, partial recovery over time, and the robustness of embeddings in speaker verification tasks.

Contribution

It is the first analysis of speaker embeddings during oral cancer treatment, highlighting their ability to capture voice changes and maintain speaker verification robustness.

Findings

Significant differences in embeddings pre- and post-treatment

Partial voice trait recovery observed after 12 months

Stable false positive rate in speaker verification

Abstract

In this paper, we analyze the behavior of speaker embeddings of patients during oral cancer treatment. First, we found that pre- and post-treatment speaker embeddings differ significantly, notifying a substantial change in voice characteristics. However, a partial recovery to pre-operative voice traits is observed after 12 months post-operation. Secondly, the same-speaker similarity at distinct treatment stages is similar to healthy speakers, indicating that the embeddings can capture characterizing features of even severely impaired speech. Finally, a speaker verification analysis signifies a stable false positive rate and variable false negative rate when combining speech samples of different treatment stages. This indicates robustness of the embeddings towards other speakers, while still capturing the changing voice characteristics during treatment. To the best of our knowledge, this is the first analysis of speaker embeddings during oral cancer treatment of patients.