CITATION.cff

cff-version: 1.2.0
title: "CASE Benchmark: Carrier-Agnostic Speaker Verification Evaluation"
message: "If you use the CASE Benchmark, please cite it as below."
type: dataset
authors:
  - family-names: "Gitter"
    given-names: "Ben"
repository-code: "https://github.com/gittb/case-benchmark"
license: CC-BY-NC-4.0
version: "1.0.0"
date-released: "2026-01-29"
keywords:
  - speaker verification
  - speaker recognition
  - benchmark
  - audio processing
  - robustness evaluation
  - carrier conditions
abstract: >-
  The CASE (Carrier-Agnostic Speaker Embedding) Benchmark evaluates speaker
  verification model robustness across real-world acoustic carrier conditions
  including codecs, microphones, noise, reverb, and playback chains. It reveals
  significant performance gaps in state-of-the-art models and demonstrates how
  carrier-aware training can improve robustness.