SPARC: Spoken Profile Authorship Recognition Corpus
We introduce SPARC (Spoken Profile Authorship Recognition Corpus), a benchmark built from two podcast corpora comprising over 12,500 speakers, 1.5M quality-filtered utterances, and speaker-level inferred or source-provided labels for sex, age, and accent, and evaluate five attribution models and three LLMs on identity- and attribute-verification trials with episode- and metadata-based topic proxies.
ProofPaper ↗
Key points
- Speech anonymization has focused on the acoustic signal, leaving open how much speaker information the transcribed words themselves retain.
- In this work, we study pairwise speaker and demographic-attribute verification from transcribed speech.
- LLM adversaries distinguish same-speaker pairs from text alone.
- Attribute-label agreement is harder to verify, but LLMs perform sex- and accent-label verification above chance.
Sources (2)
- [1]SPARC: Spoken Profile Authorship Recognition CorpusHugging Face Daily Papers · Oct 18, 12:00 AM
We introduce SPARC (Spoken Profile Authorship Recognition Corpus), a benchmark built from two podcast corpora comprising over 12,500 speakers, 1.5M quality-filtered utterances, and speaker-level inferred or source-provided labels for sex, age, and accent, and evaluate five attribution models and three LLMs on identity- and attribute-verification trials with episode- and metadata-based topic proxie
Speech anonymization has focused on the acoustic signal, leaving open how much speaker information the transcribed words themselves retain.
- [2]SPARC: Spoken Profile Authorship Recognition CorpusarXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 18, 01:34 PM · same content
Extractive summary: sentences quoted from the sources.
Before this
- Oct 17, 2026EdgeKernelBench: Unified LLM-Driven Kernel Generation for Edge Backends
- Oct 16, 2026[Research] 2,788 tokens/s on a single RTX 5090 with a diffusion LLM — Introducing ORL (fully open source)
- Oct 16, 2026Characterising and Analysing Retrieval-Augmented Generation Techniques
- Oct 16, 2026Scaling Indic Pretraining: 2.8 Trillion Tokens and an Open Data Generation Pipeline
- Oct 16, 2026OR-License: Evaluating LLM Agents as Real-World Operations Research Practitioners
- Oct 16, 2026Language Models Learn to Improve Language Models