# Graph Attention Networks for Speaker Verification

arXiv: https://arxiv.org/abs/2010.11543 · HTML: https://mlsanity.com/paper/2010.11543/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 22 Oct 2020.

## Built on by (most cited first, 6 of 6)

- [AASIST: Audio Anti-Spoofing using Integrated Spectro-Temporal Graph Attention Networks](https://mlsanity.com/paper/2110.01200/) (2021)
- [Multi-scale speaker embedding-based graph attention networks for speaker diarisation](https://mlsanity.com/paper/2110.03361/) (2021)
- [Parameter-Free Attentive Scoring for Speaker Verification](https://mlsanity.com/paper/2203.05642/) (2022)
- [Improved Relation Networks for End-to-End Speaker Verification and Identification](https://mlsanity.com/paper/2203.17218/) (2022)
- [wav2graph: A Framework for Supervised Learning Knowledge Graph from Speech](https://mlsanity.com/paper/2408.04174/) (2024)
- [Attention-based graph neural networks: a survey](https://mlsanity.com/paper/2605.08679/) (2026)

## Abstract (arXiv)

This work presents a novel back-end framework for speaker verification using graph attention networks. Segment-wise speaker embeddings extracted from multiple crops within an utterance are interpreted as node representations of a graph. The proposed framework inputs segment-wise speaker embeddings from an enrollment and a test utterance and directly outputs a similarity score. We first construct a graph using segment-wise speaker embeddings and then input these to graph attention networks. After a few graph attention layers with residual connections, each node is projected into a one-dimensional space using affine transform, followed by a readout operation resulting in a scalar similarity score. To enable successful adaptation for speaker verification, we propose techniques such as separating trainable weights for attention map calculations between segment-wise speaker embeddings from different utterances. The effectiveness of the proposed framework is validated using three different speaker embedding extractors trained with different architectures and objective functions. Experimental results demonstrate consistent improvement over various baseline back-end classifiers, with an average equal error rate improvement of 20% over the cosine similarity back-end without test time augmentation.