Stack Less, Repeat More: A Block Reusing Approach for Progressive Speech Enhancement

TL;DR

This paper introduces a novel speech enhancement method that reuses a single processing block repeatedly for progressive refinement, reducing parameters and improving efficiency.

Contribution

The proposed approach demonstrates that block reuse enables effective progressive learning in speech enhancement, reducing model complexity and redundancy.

Findings

Repeated block reuse leads to performance comparable to deeper models.

Progressive refinement occurs within a single block, reducing parameter redundancy.

Deepening encoder and decoder is unnecessary with block reuse.

Abstract

This paper presents an efficient speech enhancement (SE) approach that reuses a processing block repeatedly instead of conventional stacking. Rather than increasing the number of blocks for learning deep latent representations, repeating a single block leads to progressive refinement while reducing parameter redundancy. We also minimize domain transformation by keeping an encoder and decoder shallow and reusing a single sequence modeling block. Experimental results show that the number of processing stages is more critical to performance than the number of blocks with different weights. Also, we observed that the proposed method gradually refines a noisy input within a single block. Furthermore, with the block reuse method, we demonstrate that deepening the encoder and decoder can be redundant for learning deep complex representation. Therefore, the experimental results confirm that the proposed block reusing enables progressive learning and provides an efficient alternative for SE.