# Improved Conformer-based End-to-End Speech Recognition Using Neural Architecture Search

Source: [Improved Conformer-based End-to-End Speech Recognition Using Neural Architecture Search](<https://www.sota2.com/research/paper/68fa69bd1e209131cea23e6d>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-28T13:56:47.413Z

Authors: Yukun Liu, Ta Li, Pengyuan Zhang, Yonghong Yan

Published: 2021-04-12T12:17:55.000Z

## Abstract

Recently neural architecture search(NAS) has been successfully used in image classification, natural language processing, and automatic speech recognition(ASR) tasks for finding the state-of-the-art(SOTA) architectures than those human-designed architectures. NAS can derive a SOTA and data-specific architecture over validation data from a pre-defined search space with a search algorithm. Inspired by the success of NAS in ASR tasks, we propose a NAS-based ASR framework containing one search space and one differentiable search algorithm called Differentiable Architecture Search(DARTS). Our search space follows the convolution-augmented transformer(Conformer) backbone, which is a more expressive ASR architecture than those used in existing NAS-based ASR frameworks. To improve the performance of our method, a regulation method called Dynamic Search Schedule(DSS) is employed. On a widely used Mandarin benchmark AISHELL-1, our best-searched architecture outperforms the baseline Conform model significantly with about 11% CER relative improvement, and our method is proved to be pretty efficient by the search cost comparisons.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2104.05390>)

[PDF](<https://arxiv.org/pdf/2104.05390.pdf>)
