Voice Impression Control in Zero-Shot TTS

Research Summary

Contribution Summary

• Para-/non-linguistic information in speech is pivotal in shaping the listeners' impression.
• Although zero-shot text-to-speech (TTS) has achieved high speaker fidelity, modulating subtle para-/non-linguistic information to control perceived voice characteristics, i.e., impressions, remains challenging.
• We have therefore developed a voice impression control method in zero-shot TTS that utilizes a low-dimensional vector to represent the intensities of various voice impression pairs (e.g., dark-bright).

Why It Matters For Eval

• The results of both objective and subjective evaluations have demonstrated our method's effectiveness in impression control.

Related Papers

• When Stereotypes GTG: The Impact of Predictive Text Suggestions on Gender Bias in Human-AI Co-Writing Embedding Neighbor
• Bob's Confetti: Phonetic Memorization Attacks in Music and Video Generation Embedding Neighbor
• See and Fix the Flaws: Enabling VLMs and Diffusion Models to Comprehend Visual Artifacts via Agentic Data Synthesis Embedding Neighbor
• Modality-Guided Mixture of Graph Experts with Entropy-Triggered Routing for Multimodal Recommendation Embedding Neighbor
• Affine-Scaled Attention: Towards Flexible and Stable Transformer Attention Embedding Neighbor
• Hearing the Order: Investigating Position Bias in Large Audio-Language Models Embedding Neighbor
• The Mighty ToRR: A Benchmark for Table Reasoning and Robustness Embedding Neighbor
• Mixture-of-Experts as Soft Clustering: A Dual Jacobian-PCA Spectral Geometry Perspective Embedding Neighbor
• Provably Safe Generative Sampling with Constricting Barrier Functions Embedding Neighbor
• The Headless Firm: How AI Reshapes Enterprise Boundaries Tag Overlap
• Measuring AI Ability to Complete Long Software Tasks Tag Overlap
• Residual Connections and the Causal Shift: Uncovering a Structural Misalignment in Transformers Tag Overlap