# PCF: ECAPA-TDNN with Progressive Channel Fusion for Speaker Verification

arXiv: https://arxiv.org/abs/2303.00204 · HTML: https://mlsanity.com/paper/2303.00204/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 1 Mar 2023.

## Built on by (most cited first, 8 of 8)

- [Audio-Visual Speaker Verification via Joint Cross-Attention](https://mlsanity.com/paper/2309.16569/) (2023)
- [Text-Aware Adapter for Few-Shot Keyword Spotting](https://mlsanity.com/paper/2412.18142/) (2024)
- [DS-TDNN: Dual-stream Time-delay Neural Network with Global-aware Filter for Speaker Verification](https://mlsanity.com/paper/2303.11020/) (2023)
- [MASV: Speaker Verification with Global and Local Context Mamba](https://mlsanity.com/paper/2412.10989/) (2024)
- [Progressive Sub-Graph Clustering Algorithm for Semi-Supervised Domain Adaptation Speaker Verification](https://mlsanity.com/paper/2305.12703/) (2023)
- [LightCAM: A Fast and Light Implementation of Context-Aware Masking based D-TDNN for Speaker Verification](https://mlsanity.com/paper/2402.06073/) (2024)
- [GMM-ResNext: Combining Generative and Discriminative Models for Speaker Verification](https://mlsanity.com/paper/2407.03135/) (2024)
- [MECT: Mixture of Experts with CNN-Transformer Network for Speaker verification](https://mlsanity.com/paper/2609.24061/) (2026)

## Abstract (arXiv)

ECAPA-TDNN is currently the most popular TDNN-series model for speaker verification, which refreshed the state-of-the-art(SOTA) performance of TDNN models. However, one-dimensional convolution has a global receptive field over the feature channel. It destroys the time-frequency relevance of the spectrogram. Besides, as ECAPA-TDNN only has five layers, a much shallower structure compared to ResNet restricts the capability to generate deep representations. To further improve ECAPA-TDNN, we propose a progressive channel fusion strategy that splits the spectrogram across the feature channel and gradually expands the receptive field through the network. Secondly, we enlarge the model by extending the depth and adding branches. Our proposed model achieves EER with 0.718 and minDCF(0.01) with 0.0858 on vox1o, relatively improved 16.1% and 19.5% compared with ECAPA-TDNN-large.