An Empirical Study on the Impact of Positional Encoding in Transformer-based Monaural Speech Enhancement

TL;DR

This study empirically evaluates how different positional encoding methods affect Transformer-based speech enhancement, revealing that positional encoding benefits noncausal models but is less helpful for causal models, with relative embeddings outperforming absolute ones.

Contribution

It provides a comprehensive comparison of five positional encoding methods in Transformer speech enhancement, clarifying their impact in causal and noncausal settings.

Findings

Positional encoding is less helpful for causal models.

Models benefit from positional encoding in noncausal configurations.

Relative position embeddings outperform absolute position embeddings.

Abstract

Transformer architecture has enabled recent progress in speech enhancement. Since Transformers are position-agostic, positional encoding is the de facto standard component used to enable Transformers to distinguish the order of elements in a sequence. However, it remains unclear how positional encoding exactly impacts speech enhancement based on Transformer architectures. In this paper, we perform a comprehensive empirical study evaluating five positional encoding methods, i.e., Sinusoidal and learned absolute position embedding (APE), T5-RPE, KERPLE, as well as the Transformer without positional encoding (No-Pos), across both causal and noncausal configurations. We conduct extensive speech enhancement experiments, involving spectral mapping and masking methods. Our findings establish that positional encoding is not quite helpful for the models in a causal configuration, which indicates that causal attention may implicitly incorporate position information. In a noncausal configuration, the models significantly benefit from the use of positional encoding. In addition, we find that among the four position embeddings, relative position embeddings outperform APEs.