On Feature Importance and Interpretability of Speaker Representations
원문 초록
Unsupervised speech disentanglement aims at separating fast varying from slowly varying components of a speech signal. In this contribution, we take a closer look at the embedding vector representing the slowly varying signal components, commonly named the speaker embedding vector. We ask, which properties of a speaker's voice are captured and investigate to which extent do individual embedding vector components sign responsible for them, using the concept of Shapley values. Our findings show that certain speaker-specific acoustic-phonetic properties can be fairly well predicted from the speaker embedding, while the investigated more abstract voice quality features cannot.
함께 살펴볼 논문
공통 태그를 근거로 연결했습니다. 실제 인용 관계나 논문의 우수성을 뜻하지 않습니다.
Adversarial Disentanglement by Backpropagation with Physics-Informed Variational Autoencoder
Enhancing interpretability of rule-based classifiers through feature graphs
Dual-View Disentangled Multi-Intent Learning for Enhanced Collaborative Filtering
Profiling Electric Vehicles via Early Charging Voltage Patterns
Instruction-Tuned Video-Audio Models Elucidate Functional Specialization in the Brain
The Reader is the Metric: How Textual Features and Reader Profiles Explain Conflicting Evaluations of AI Creative Writing
토론
이 논문과 연결된 포럼 글이 아직 없습니다.