ADVOSYNTH: A Synthetic Multi-Advocate Dataset for Speaker Identification in Courtroom Scenarios

TL;DR

Advosynth-500 introduces a courtroom-focused speaker identification benchmark generated with Speech Llama Omni to study synthetic voices under adversarial dialogue. The dataset defines 10 advocate identities via a Vocal Identity Matrix $V_i = \{P_i, R_i, T_i, \sigma_i\}$ and yields a structured, multi-argument corpus that the authors use to test closed-set SID and voice constancy under varying prosody. A core contribution is the combination of end-to-end speech-to-speech synthesis with domain-specific vocal dynamics to create a realistic, labeled evaluation ground truth for forensic SID tools, including potentially state-of-the-art embeddings like X-vectors and ECAPA-TDNN. The work emphasizes the privacy-preserving, auditable nature of synthetic data for forensic evaluation and sets a new standard for stress-testing audio evidence tools in judicial contexts.

Abstract

As large-scale speech-to-speech models achieve high fidelity, the distinction between synthetic voices in structured environments becomes a vital area of study. This paper introduces Advosynth-500, a specialized dataset comprising 100 synthetic speech files featuring 10 unique advocate identities. Using the Speech Llama Omni model, we simulate five distinct advocate pairs engaged in courtroom arguments. We define specific vocal characteristics for each advocate and present a speaker identification challenge to evaluate the ability of modern systems to map audio files to their respective synthetic origins. Dataset is available at this link-https: //github.com/naturenurtureelite/ADVOSYNTH-500.