HydraFormer: One Encoder For All Subsampling Rates
This addresses cost inefficiencies for speech recognition practitioners by enabling a single model to adapt to diverse scenarios, though it is incremental as it builds on existing Conformer and Transformer architectures.
The paper tackles the problem of needing multiple models for different subsampling rates in automatic speech recognition by proposing HydraFormer, which uses a multi-branch encoder to flexibly handle various rates, reducing training and deployment costs while maintaining high recognition performance on datasets like AISHELL-1 and LibriSpeech.
In automatic speech recognition, subsampling is essential for tackling diverse scenarios. However, the inadequacy of a single subsampling rate to address various real-world situations often necessitates training and deploying multiple models, consequently increasing associated costs. To address this issue, we propose HydraFormer, comprising HydraSub, a Conformer-based encoder, and a BiTransformer-based decoder. HydraSub encompasses multiple branches, each representing a distinct subsampling rate, allowing for the flexible selection of any branch during inference based on the specific use case. HydraFormer can efficiently manage different subsampling rates, significantly reducing training and deployment expenses. Experiments on AISHELL-1 and LibriSpeech datasets reveal that HydraFormer effectively adapts to various subsampling rates and languages while maintaining high recognition performance. Additionally, HydraFormer showcases exceptional stability, sustaining consistent performance under various initialization conditions, and exhibits robust transferability by learning from pretrained single subsampling rate automatic speech recognition models\footnote{Model code and scripts: https://github.com/HydraFormer/hydraformer}.