Playing a Part: Speaker Verification at the Movies

TL;DR

This paper evaluates how well current speaker recognition models perform on movie speech data with disguises and domain differences, introduces a new challenging dataset, and explores domain adaptation techniques.

Contribution

Introduces VoxMovies, a new challenging dataset for speaker recognition in movies, and benchmarks model performance with domain adaptation methods.

Findings

Model performance drops significantly on movie data.

Domain adaptation improves accuracy but leaves room for improvement.

VoxMovies dataset highlights challenges in real-world speaker recognition.

Abstract

The goal of this work is to investigate the performance of popular speaker recognition models on speech segments from movies, where often actors intentionally disguise their voice to play a character. We make the following three contributions: (i) We collect a novel, challenging speaker recognition dataset called VoxMovies, with speech for 856 identities from almost 4000 movie clips. VoxMovies contains utterances with varying emotion, accents and background noise, and therefore comprises an entirely different domain to the interview-style, emotionally calm utterances in current speaker recognition datasets such as VoxCeleb; (ii) We provide a number of domain adaptation evaluation sets, and benchmark the performance of state-of-the-art speaker recognition models on these evaluation pairs. We demonstrate that both speaker verification and identification performance drops steeply on this new data, showing the challenge in transferring models across domains; and finally (iii) We show that simple domain adaptation paradigms improve performance, but there is still large room for improvement.