关键词: 关键词: 声学特征, 多路信息聚合, 双路编码器, 三路信息聚合解码器, 通道-时频注意力机制

Abstract: In order to address the issues of insufficient acoustic feature extraction and severe decoding feature loss in single-channel speech enhancement networks based on convolutional encoder-decoder architecture, this paper proposes a Monaural speech enhancement network called Multi-Channel Information Aggregation and Collaborative Decoding (MIACD). The MIACD network utilizes a dual-channel encoder to extract the speech magnitude spectrum and complex spectrum features, which are enriched with self-supervised learning (SSL) representations. It employs a four-layer Conformer block to model the extracted features in both the time and frequency domains. By incorporating residual connections, the speech magnitude and complex features extracted by the dual-channel encoder are introduced into a three-channel information aggregation decoder. Additionally, a channel-time-frequency attention mechanism (CTF-Attention) is proposed to adjust the aggregated information in the decoder based on the distribution of speech energy, effectively improving the problem of severe acoustic information loss during decoding. Experimental results on the publicly available dataset Voice Bank DEMAND demonstrate that, compared to Glance and Gaze: A Collaborative Learning Framework for Single-channel Speech Enhancement (GaGNet), the proposed method achieves a 5.1% improvement on the objective metric WB-PESQ and 96.7 on STOI, validating that the proposed method effectively utilizes speech information for signal reconstruction, noise suppression, and speech intelligibility enhancement.

Key words: Keywords: acoustic features, multi-channel information aggregation, dual-channel encoder, three-channel information Aggregation decoder, channel-time-frequency attention mechanism