Conv-TasNet variant improves speech separation with less computing power
EConv-TasNet: Efficient Conv-TasNet for Effective Speech Separation
Sound
Summary
Separating different voices in a recording is often done by a method called Conv-TasNet, but many improved versions need lots of computing power. The authors designed a new version called eConv-TasNet that works faster and uses less memory without needing complicated parts. It uses two new modules to better identify speakers and combine their features during processing. Tests show it improves voice separation quality while running more efficiently, making it better suited for use in devices with limited resources.
What this means in practice
- •For mobile app developers: Build voice separation features in apps that run smoothly on smartphones with limited processing power.
- •For embedded systems engineers: Integrate efficient speech separation into hardware devices like smart speakers or hearing aids with strict resource constraints.
Authors
Pei-Chun Chang, Chuan-Yi Liu
Abstract
Conv-TasNet has served as a strong baseline for time-domain speech separation, and many studies have extended it with advanced architectures such as dual-path networks, U-Nets, and attention mechanisms. However, these methods often introduce high computational cost and complexity, limiting their deployment in resource-constrained scenarios. To address this issue, we propose eConv-TasNet, an efficient variant of Conv-TasNet that improves both effectiveness and efficiency without relying on resource-intensive modules. The proposed model consists of a group-wise early-splitting (GES) module and a multi-group feature aggregation (MGFA) module. GES generates discriminative speaker embeddings at intermediate stages, while MGFA progressively aggregates these group-level representations for refined mask estimation. Experimental results show that eConv-TasNet reduces model size by 22.4%, accelerates inference by 18.9%, and improves SI-SNRi by 14.0%-28.0% across three public benchmarks. Moreover, it achieves competitive performance compared with state-of-the-art methods while requiring significantly fewer parameters and lower inference cost. These results demonstrate a favorable efficiency-effectiveness trade-off for edge deployment.