LGMLMar 9, 2018

Scoring Formulation for Multi-Condition Joint PLDA

arXiv:1803.03684v1
Originality Synthesis-oriented
AI Analysis

This work is incremental, providing a technical extension for researchers in speaker recognition or similar fields using PLDA models.

The paper tackles the problem of scoring in joint PLDA models with multiple nuisance conditions, extending the original single-condition formulation to handle multiple conditions and deriving the necessary likelihood ratios for scoring.

The joint PLDA model, is a generalization of PLDA where the nuisance variable is no longer considered independent across samples, but potentially shared (tied) across samples that correspond to the same nuisance condition. The original work considered a single nuisance condition, deriving the EM and scoring formulas for this scenario. In this document, we show how to obtain likelihood ratios for scoring when multiple nuisance conditions are allowed in the model.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes