MERaLiON-GR: Speech Gender Recognition Model for English and SEA Languages
Quick Answer
MERaLiON-GR is a speech gender recognition model that outperforms Vox-Profile and a large Audio-LLM in classifying gender across multiple languages, including English and Southeast Asian languages.
Quick Take
Utilizing a fine-tuned MERaLiON-SpeechEncoder-2 and a multi-scale ECAPA-TDNN network, it demonstrates superior performance in both full-utterance and segment-level evaluations, highlighting the importance of specialized speech models for accurate paralinguistic understanding.
Key Points
- MERaLiON-GR performs binary gender classification for English and SEA languages.
- The model fine-tunes MERaLiON-SpeechEncoder-2 using Low-Rank Adaptation (LoRA).
- It surpasses Vox-Profile and a large Audio- in multilingual evaluations.
- Results show strong cross-lingual generalization and accurate paralinguistic understanding.
- Evaluated languages include English, Chinese, Malay, Tamil, Thai, Vietnamese, Indonesian, and Khmer.
DeepSignal Analysis
What happened
The MERaLiON-GR model has been developed for speech gender recognition, achieving superior performance in classifying gender in English and several Southeast Asian languages. It utilizes a fine-tuned speech encoder and a multi-scale network for effective gender classification.
Key evidence
- MERaLiON-GR performs binary classification of gender (female/male) across multiple languages, including English and Southeast Asian languages.
- The model employs a fine-tuned MERaLiON-SpeechEncoder-2 and a multi-scale ECAPA-TDNN network, demonstrating its effectiveness in both full-utterance and segment-level evaluations.
- In evaluations, MERaLiON-GR consistently outperforms the Vox-Profile model and a large Audio-LLM in gender recognition tasks.
Why it matters
The development of MERaLiON-GR highlights the significance of specialized models in speech processing, particularly for tasks like gender recognition. Its performance across multiple languages indicates potential applications in diverse linguistic contexts, which can enhance user experience in voice-activated systems and improve accessibility.
What to watch
Paper Resources
📖 Reader Mode
~2 min readAuthors:Qiongqiong Wang, Ai Ti Aw, Nancy F. Chen, Ying Lay Chiu, Yang Ding, Yingxu He, Ridong Jiang, Zhuohan Liu, Yanfeng Lu, Yi Ma, Muhammad Huzaifah, Nabilah Binte Md Johan, Nattadaporn Lertcheva, Pham Minh Duc, Sailor Hardik Bhupendra, Siti Umairah Binte Mohammad Salleh, Shuo Sun, Tarun Kumar Vangani, Jeremy H. M. Wong, Jinyang Wu, Longyin Zhang
Abstract:We present MERaLiON-GR, a speech gender recognition system that performs binary classification (female / male) on English and Southeast Asian (SEA) languages. The model finetunes MERaLiON-SpeechEncoder-2, a large conformer based transformer pre-trained on a broad speech corpus, and applies parameter efficient fine-tuning via Low-Rank Adaptation (LoRA) to adapt the encoder to the gender recognition task, and appends a multi-scale ECAPA-TDNN down stream network with attention pooling and a lightweight linear classifier. Extensive evaluations across multilingual Singaporean and Southeast Asian languages (English, Chinese, Malay, Tamil, Thai, Vietnamese, Indonesian, and Khmer) show that MERaLiON-GR consistently surpasses the state-of-the-art gender recognition model Vox-Profile and a large Audio-LLM, in both full-utterance and segment level evaluation modes. The results underscore the value of dedicated speech models in achieving accurate paralinguistic understanding and strong cross-lingual generalization.
| Subjects: | Computation and Language (cs.CL); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2608.04433 [cs.CL] |
| (or arXiv:2608.04433v1 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2608.04433 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Qiongqiong Wang [view email]
[v1]
Wed, 5 Aug 2026 04:22:27 UTC (21 KB)
— Originally published at arxiv.org
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from arXiv cs.CL
See more →TriAgent: Divergence-Aware Committees for Cost-Efficient Financial Sentiment Analysis
TriAgent introduces a cost-efficient multi-agent system for financial sentiment analysis, combining VADER, FinBERT, and Qwen2.5. It achieves an F1 score of ~0.87 with significant savings of $9.3M/year at a 10M-user scale compared to GPT-4o-mini, while also detecting hallucinations with an AUC of 0.90.