⌕ Поиск по атласу

NIST · ИТ · bbef864e6fdff146bd0a

bbef864e6fdff146bd0a · Beyond EER: Multi-Dimensional Evaluation of Information Leakage in Speaker De-Identification

Beyond EER: Multi-Dimensional Evaluation of Information Leakage in Speaker De-Identification

Speaker de-identification (SDID) aims to preserve privacy by concealing speaker identity while maintaining speech utility. However, current evaluations often reduce privacy to a single dimension—biometric verification performance—typically measured by

Официальный источник · 2026-09-01

Полные импортированные сведения карточки
{
  "title_en": "Beyond EER: Multi-Dimensional Evaluation of Information Leakage in Speaker De-Identification",
  "designation": "",
  "series": "",
  "status": "",
  "date": "2026-09-01",
  "date_original": "September 1, 2026",
  "url": "https://www.nist.gov/publications/beyond-eer-multi-dimensional-evaluation-information-leakage-speaker-de-identification",
  "source": "NIST Publications",
  "display_en": "",
  "authors_en": "Seungmin Seo , Oleg Aulov , P Phillips , Kevin Mangold , Jonathan Eskin",
  "abstract_excerpt_en": "Speaker de-identification (SDID) aims to preserve privacy by concealing speaker identity while maintaining speech utility. However, current evaluations often reduce privacy to a single dimension—biometric verification performance—typically measured by",
  "id": "bbef864e6fdff146bd0a",
  "sources": [
    "NIST Publications"
  ],
  "source_records": [
    {
      "title_en": "Beyond EER: Multi-Dimensional Evaluation of Information Leakage in Speaker De-Identification",
      "designation": "",
      "series": "",
      "status": "",
      "date": "2026-09-01",
      "date_original": "September 1, 2026",
      "url": "https://www.nist.gov/publications/beyond-eer-multi-dimensional-evaluation-information-leakage-speaker-de-identification",
      "source": "NIST Publications",
      "display_en": "",
      "authors_en": "Seungmin Seo , Oleg Aulov , P Phillips , Kevin Mangold , Jonathan Eskin",
      "abstract_excerpt_en": "Speaker de-identification (SDID) aims to preserve privacy by concealing speaker identity while maintaining speech utility. However, current evaluations often reduce privacy to a single dimension—biometric verification performance—typically measured by"
    }
  ],
  "translation_state": "none"
}