Personalized Speech Enhancement Combining Band-split RNN And Speaker Attentive Module

Abstract

Target speaker information can be utilized in speech enhancement (SE) models to more effectively extract the desired speech. Previous works introduce the speaker embedding into speech enhancement models by means of concatenation or affine transformation. In this paper, we propose a speaker attentive module to calculate the attention scores between the speaker embedding and the intermediate features, which are used to rescale the features. By merging this module in the state-of-the-art SE model, we construct the personalized SE model for ICASSP Signal Processing Grand Challenge: DNS Challenge 5 (2023). Our system achieves a final score of 0.529 on the blind test set of track1 and 0.549 on track2.

Stats

• citations0
• S2 citations—
• github stars0
• HF likes0
• heat score0.00
• arxiv keyle2023personalized

Related papers

• High Fidelity Speech Enhancement With Band-split RNN (2022)0.00
• Personalized Speech Enhancement Without A Separate Speaker Embedding Model (2024)5.24
• Parallel Gated Neural Network With Attention Mechanism For Speech Enhancement (2022)0.00
• Sef-pnet: Speaker Encoder-free Personalized Speech Enhancement With Local And Global Contexts Aggregation (2025)2.26
• Personalized Percepnet: Real-time, Low-complexity Target Voice Separation And Enhancement (2021)10.97
• Multi-frequency Information Enhanced Channel Attention Module For Speaker Representation Learning (2022)0.00
• Zero-shot Personalized Speech Enhancement Through Speaker-informed Model Selection (2021)7.16
• Bridging The Gap: Integrating Pre-trained Speech Enhancement And Recognition Models For Robust Speech Recognition (2024)7.50