# Graph-based Label Propagation for Semi-Supervised Speaker Identification

arXiv: https://arxiv.org/abs/2106.08207 · HTML: https://mlsanity.com/paper/2106.08207/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- Builds on 18 of its 39 references that are in the map.
- First posted on arXiv 15 Jun 2021.

## Builds on (most cited first, 18 of 18)

- [Semi-Supervised Classification with Graph Convolutional Networks](https://mlsanity.com/paper/1609.02907/) (2016)
- [Simplifying Graph Convolutional Networks](https://mlsanity.com/paper/1902.07153/) (2019)
- [Temporal Ensembling for Semi-Supervised Learning](https://mlsanity.com/paper/1610.02242/) (2016)
- [Predict then Propagate: Graph Neural Networks meet Personalized PageRank](https://mlsanity.com/paper/1810.05997/) (2018)
- [Virtual Adversarial Training: A Regularization Method for Supervised and Semi-Supervised Learning](https://mlsanity.com/paper/1704.03976/) (2017)
- [VoxCeleb2: Deep Speaker Recognition](https://mlsanity.com/paper/1806.05622/) (2018)
- [VoxCeleb: a large-scale speaker identification dataset](https://mlsanity.com/paper/1706.08612/) (2017)
- [Generalizing from a Few Examples: A Survey on Few-Shot Learning](https://mlsanity.com/paper/1904.05046/) (2019)
- [Deep Learning on Graphs: A Survey](https://mlsanity.com/paper/1812.04202/) (2018)
- [Generalized End-to-End Loss for Speaker Verification](https://mlsanity.com/paper/1710.10467/) (2017)
- [Combining Label Propagation and Simple Models Out-performs Graph Neural Networks](https://mlsanity.com/paper/2010.13993/) (2020)
- [Deep Metric Transfer for Label Propagation with Limited Annotated Data](https://mlsanity.com/paper/1812.08781/) (2018)
- [Semi-Supervised Contrastive Learning with Generalized Contrastive Loss and Its Application to Speaker Recognition](https://mlsanity.com/paper/2006.04326/) (2020)
- [Attention-Based Models for Text-Dependent Speaker Verification](https://mlsanity.com/paper/1710.10470/) (2017)
- [AutoSpeech: Neural Architecture Search for Speaker Recognition](https://mlsanity.com/paper/2005.03215/) (2020)
- [Distributed Deep Learning Strategies For Automatic Speech Recognition](https://mlsanity.com/paper/1904.04956/) (2019)
- [Cosine-Distance Virtual Adversarial Training for Semi-Supervised Speaker-Discriminative Acoustic Embeddings](https://mlsanity.com/paper/2008.03756/) (2020)
- [Speaker attribution with voice profiles by graph-based semi-supervised learning](https://mlsanity.com/paper/2102.03634/) (2021)

## Built on by (most cited first, 4 of 4)

- [Efficient Graph Optimization via Distance-Aware Graph Representation Learning](https://mlsanity.com/paper/2406.17281/) (2024)
- [Graph-based Multi-View Fusion and Local Adaptation: Mitigating Within-Household Confusability for Speaker Identification](https://mlsanity.com/paper/2207.04081/) (2022)
- [Cross-utterance ASR Rescoring with Graph-based Label Propagation](https://mlsanity.com/paper/2303.15132/) (2023)
- [FewShotTextGCN: K-hop neighborhood regularization for few-shot learning on graphs](https://mlsanity.com/paper/2301.10481/) (2023)

## Abstract (arXiv)

Speaker identification in the household scenario (e.g., for smart speakers) is typically based on only a few enrollment utterances but a much larger set of unlabeled data, suggesting semisupervised learning to improve speaker profiles. We propose a graph-based semi-supervised learning approach for speaker identification in the household scenario, to leverage the unlabeled speech samples. In contrast to most of the works in speaker recognition that focus on speaker-discriminative embeddings, this work focuses on speaker label inference (scoring). Given a pre-trained embedding extractor, graph-based learning allows us to integrate information about both labeled and unlabeled utterances. Considering each utterance as a graph node, we represent pairwise utterance similarity scores as edge weights. Graphs are constructed per household, and speaker identities are propagated to unlabeled nodes to optimize a global consistency criterion. We show in experiments on the VoxCeleb dataset that this approach makes effective use of unlabeled data and improves speaker identification accuracy compared to two state-of-the-art scoring methods as well as their semi-supervised variants based on pseudo-labels.