SDCLASMay 21, 2025

MIKU-PAL: An Automated and Standardized Multi-Modal Method for Speech Paralinguistic and Affect Labeling

arXiv:2505.15772v14 citationsh-index: 6Proceedings of the 1st Joint Workshop on Large Language Models and Structure Modeling (XLLM 2025)
Originality Incremental advance
AI Analysis

This work addresses the problem of costly and inconsistent emotional speech annotation for speech synthesis researchers, offering a cheaper and faster automated solution.

The paper tackles the challenge of acquiring large-scale emotional speech data with strong consistency for speech synthesis by presenting MIKU-PAL, an automated multimodal pipeline that extracts high-consistency emotional speech from unlabeled video data, achieving human-level accuracy (68.5% on MELD) and superior consistency (0.93 Fleiss kappa score).

Acquiring large-scale emotional speech data with strong consistency remains a challenge for speech synthesis. This paper presents MIKU-PAL, a fully automated multimodal pipeline for extracting high-consistency emotional speech from unlabeled video data. Leveraging face detection and tracking algorithms, we developed an automatic emotion analysis system using a multimodal large language model (MLLM). Our results demonstrate that MIKU-PAL can achieve human-level accuracy (68.5% on MELD) and superior consistency (0.93 Fleiss kappa score) while being much cheaper and faster than human annotation. With the high-quality, flexible, and consistent annotation from MIKU-PAL, we can annotate fine-grained speech emotion categories of up to 26 types, validated by human annotators with 83% rationality ratings. Based on our proposed system, we further released a fine-grained emotional speech dataset MIKU-EmoBench(131.2 hours) as a new benchmark for emotional text-to-speech and visual voice cloning.

Foundations

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

Your Notes