MECT: Mixture of Experts with CNN-Transformer Network for Speaker verification

2026-09-20

AI Summary

本文提出MECT模型，通过在CNN-Transformer框架中引入混合专家机制以优化说话人验证性能，同时保持参数量和计算复杂度较低。

Abstract

In this paper, we propose MECT, a speaker verification model that integrates the Mixture-of-Experts (MoE) mechanism into a CNN-Transformer backbone with optimized block structure and stacking scheme. Specifically, we investigated four MoE variants that span utterance-level and frame-level granularity with dense and sparse routing strategies. The MoE mechanism proves to be effective over the baseline without MoE with only a small increase in parameters. We further scale MECT to a series of model sizes, all maintaining compact parameters and low computational complexity. In particular, MECT-B2 achieves state-of-the-art performance on VoxCeleb1 and delivers strong results on CN-Celeb, demonstrating its effectiveness across diverse datasets. In addition, we establish a streaming inference paradigm through causal retraining, which maintains strong performance at a chunk size of 100ms.

Problem

• Speaker verification
• Mixture-of-Experts
• CNN-Transformer
• Parameter efficiency
• Computational complexity

Innovation

• Mixture-of-Experts (MoE)
• CNN-Transformer
• speaker verification
• streaming inference
• causal retraining