# SpeechNAS: Towards Better Trade-off between Latency and Accuracy for Large-Scale Speaker Verification

arXiv: https://arxiv.org/abs/2109.08839 · HTML: https://mlsanity.com/paper/2109.08839/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 26 of its 54 references that are in the map.
- First posted on arXiv 18 Sep 2021.

## Builds on (most cited first, 26 of 26)

- [Squeeze-and-Excitation Networks](https://mlsanity.com/paper/1709.01507/) (2017)
- [DARTS: Differentiable Architecture Search](https://mlsanity.com/paper/1806.09055/) (2018)
- [SpecAugment: A Simple Data Augmentation Method for Automatic Speech Recognition](https://mlsanity.com/paper/1904.08779/) (2019)
- [VoxCeleb2: Deep Speaker Recognition](https://mlsanity.com/paper/1806.05622/) (2018)
- [Once-for-All: Train One Network and Specialize it for Efficient Deployment](https://mlsanity.com/paper/1908.09791/) (2019)
- [VoxCeleb: a large-scale speaker identification dataset](https://mlsanity.com/paper/1706.08612/) (2017)
- [ECAPA-TDNN: Emphasized Channel Attention, Propagation and Aggregation in TDNN Based Speaker Verification](https://mlsanity.com/paper/2005.07143/) (2020)
- [Res2Net: A New Multi-scale Backbone Architecture](https://mlsanity.com/paper/1904.01169/) (2019)
- [MUSAN: A Music, Speech, and Noise Corpus](https://mlsanity.com/paper/1510.08484/) (2015)
- [Single Path One-Shot Neural Architecture Search with Uniform Sampling](https://mlsanity.com/paper/1904.00420/) (2019)
- [Dilated Residual Networks](https://mlsanity.com/paper/1705.09914/) (2017)
- [Generalized End-to-End Loss for Speaker Verification](https://mlsanity.com/paper/1710.10467/) (2017)
- [In defence of metric learning for speaker recognition](https://mlsanity.com/paper/2003.11982/) (2020)
- [Learning towards Minimum Hyperspherical Energy](https://mlsanity.com/paper/1805.09298/) (2018)
- [Co-occurrence Feature Learning for Skeleton based Action Recognition using Regularized Deep LSTM Networks](https://mlsanity.com/paper/1603.07772/) (2016)
- [Utterance-level Aggregation For Speaker Recognition In The Wild](https://mlsanity.com/paper/1902.10107/) (2019)
- [OpenBox: A Generalized Black-box Optimization Service](https://mlsanity.com/paper/2106.00421/) (2021)
- [Large Margin Softmax Loss for Speaker Verification](https://mlsanity.com/paper/1904.03479/) (2019)
- [Cycle-Consistent Adversarial Autoencoders for Unsupervised Text Style Transfer](https://mlsanity.com/paper/2010.00735/) (2020)
- [AutoSpeech: Neural Architecture Search for Speaker Recognition](https://mlsanity.com/paper/2005.03215/) (2020)
- [Evolutionary Algorithm Enhanced Neural Architecture Search for Text-Independent Speaker Verification](https://mlsanity.com/paper/2008.05695/) (2020)
- [Short utterance compensation in speaker verification via cosine-based teacher-student learning of speaker embeddings](https://mlsanity.com/paper/1810.10884/) (2018)
- [Embeddings for DNN speaker adaptive training](https://mlsanity.com/paper/1909.13537/) (2019)
- [Self-Adaptive Soft Voice Activity Detection using Deep Neural Networks for Robust Speaker Verification](https://mlsanity.com/paper/1909.11886/) (2019)
- [Text-Independent Speaker Verification with Dual Attention Network](https://mlsanity.com/paper/2009.05485/) (2020)
- [Adapting End-to-End Neural Speaker Verification to New Languages and Recording Conditions with Adversarial Training](https://mlsanity.com/paper/1811.03055/) (2018)

## Built on by (most cited first, 2 of 2)

- [AutoShot: A Short Video Dataset and State-of-the-Art Shot Boundary Detection](https://mlsanity.com/paper/2304.06116/) (2023)
- [Multiscale Audio Spectrogram Transformer for Efficient Audio Classification](https://mlsanity.com/paper/2303.10757/) (2023)

## Abstract (arXiv)

Recently, x-vector has been a successful and popular approach for speaker verification, which employs a time delay neural network (TDNN) and statistics pooling to extract speaker characterizing embedding from variable-length utterances. Improvement upon the x-vector has been an active research area, and enormous neural networks have been elaborately designed based on the x-vector, eg, extended TDNN (E-TDNN), factorized TDNN (F-TDNN), and densely connected TDNN (D-TDNN). In this work, we try to identify the optimal architectures from a TDNN based search space employing neural architecture search (NAS), named SpeechNAS. Leveraging the recent advances in the speaker recognition, such as high-order statistics pooling, multi-branch mechanism, D-TDNN and angular additive margin softmax (AAM) loss with a minimum hyper-spherical energy (MHE), SpeechNAS automatically discovers five network architectures, from SpeechNAS-1 to SpeechNAS-5, of various numbers of parameters and GFLOPs on the large-scale text-independent speaker recognition dataset VoxCeleb1. Our derived best neural network achieves an equal error rate (EER) of 1.02% on the standard test set of VoxCeleb1, which surpasses previous TDNN based state-of-the-art approaches by a large margin. Code and trained weights are in https://github.com/wentaozhu/speechnas.git