Multiformer: A Head-Configurable Transformer-Based Model for Direct Speech Translation

TL;DR

The paper introduces Multiformer, a transformer model with configurable attention heads for direct speech translation, improving performance by diversifying token interactions and reducing information loss.

Contribution

It proposes a novel head-configurable attention mechanism in transformers, enhancing speech translation accuracy over traditional models.

Findings

Multiformer outperforms baseline by up to 0.7 BLEU.

Diverse attention heads improve token interaction extraction.

Uniform head relevance correlates with better results.

Abstract

Transformer-based models have been achieving state-of-the-art results in several fields of Natural Language Processing. However, its direct application to speech tasks is not trivial. The nature of this sequences carries problems such as long sequence lengths and redundancy between adjacent tokens. Therefore, we believe that regular self-attention mechanism might not be well suited for it. Different approaches have been proposed to overcome these problems, such as the use of efficient attention mechanisms. However, the use of these methods usually comes with a cost, which is a performance reduction caused by information loss. In this study, we present the Multiformer, a Transformer-based model which allows the use of different attention mechanisms on each head. By doing this, the model is able to bias the self-attention towards the extraction of more diverse token interactions, and the information loss is reduced. Finally, we perform an analysis of the head contributions, and we observe that those architectures where all heads relevance is uniformly distributed obtain better results. Our results show that mixing attention patterns along the different heads and layers outperforms our baseline by up to 0.7 BLEU.