SICRN: Advancing Speech Enhancement through State Space Model and Inplace Convolution Techniques

Speech enhancement aims to improve speech quality and intelligibility, especially in noisy environments where background noise degrades speech signals. Currently, deep learning methods achieve great success in speech enhancement, e.g. the representative convolutional recurrent neural network (CRN) and its variants. However, CRN typically employs consecutive downsampling and upsampling convolution for frequency modeling, which destroys the inherent structure of the signal over frequency. Additionally, convolutional layers lacks of temporal modelling abilities. To address these issues, we propose an innovative module combing a State space model and Inplace Convolution (SIC), and to replace the conventional convolution in CRN, called SICRN. Specifically, a dual-path multidimensional State space model captures the global frequencies dependency and long-term temporal dependencies. Meanwhile, the 2D-inplace convolution is used to capture the local structure, which abandons the downsampling and upsampling. Systematic evaluations on the public INTERSPEECH 2020 DNS challenge dataset demonstrate SICRN's efficacy. Compared to strong baselines, SICRN achieves performance close to state-of-the-art while having advantages in model parameters, computations, and algorithmic delay. The proposed SICRN shows great promise for improved speech enhancement.

3S-TSE: Efficient Three-Stage Target Speaker Extraction for Real-Time and Low-Resource Applications

Target speaker extraction (TSE) aims to isolate a specific voice from multiple mixed speakers relying on a registerd sample. Since voiceprint features usually vary greatly, current end-to-end neural networks require large model parameters which are computational intensive and impractical for real-time applications, espetially on resource-constrained platforms. In this paper, we address the TSE task using microphone array and introduce a novel three-stage solution that systematically decouples the process: First, a neural network is trained to estimate the direction of the target speaker. Second, with the direction determined, the Generalized Sidelobe Canceller (GSC) is used to extract the target speech. Third, an Inplace Convolutional Recurrent Neural Network (ICRN) acts as a denoising post-processor, refining the GSC output to yield the final separated speech. Our approach delivers superior performance while drastically reducing computational load, setting a new standard for efficient real-time target speaker extraction.

Hierarchical Modeling of Spatial Cues via Spherical Harmonics for Multi-Channel Speech Enhancement

Multi-channel speech enhancement utilizes spatial information from multiple microphones to extract the target speech. However, most existing methods do not explicitly model spatial cues, instead relying on implicit learning from multi-channel spectra. To better leverage spatial information, we propose explicitly incorporating spatial modeling by applying spherical harmonic transforms (SHT) to the multi-channel input. In detail, a hierarchical framework is introduced whereby lower order harmonics capturing broader spatial patterns are estimated first, then combined with higher orders to recursively predict finer spatial details. Experiments on TIMIT demonstrate the proposed method can effectively recover target spatial patterns and achieve improved performance over baseline models, using fewer parameters and computations. Explicitly modeling spatial information hierarchically enables more effective multi-channel speech enhancement.

PDPCRN: Parallel Dual-Path CRN with Bi-directional Inter-Branch Interactions for Multi-Channel Speech Enhancement

Multi-channel speech enhancement seeks to utilize spatial information to distinguish target speech from interfering signals. While deep learning approaches like the dual-path convolutional recurrent network (DPCRN) have made strides, challenges persist in effectively modeling inter-channel correlations and amalgamating multi-level information. In response, we introduce the Parallel Dual-Path Convolutional Recurrent Network (PDPCRN). This acoustic modeling architecture has two key innovations. First, a parallel design with separate branches extracts complementary features. Second, bi-directional modules enable cross-branch communication. Together, these facilitate diverse representation fusion and enhanced modeling. Experimental validation on TIMIT datasets underscores the prowess of PDPCRN. Notably, against baseline models like the standard DPCRN, PDPCRN not only outperforms in PESQ and STOI metrics but also boasts a leaner computational footprint with reduced parameters.

MC-SpEx: Towards Effective Speaker Extraction with Multi-Scale Interfusion and Conditional Speaker Modulation

The previous SpEx+ has yielded outstanding performance in speaker extraction and attracted much attention. However, it still encounters inadequate utilization of multi-scale information and speaker embedding. To this end, this paper proposes a new effective speaker extraction system with multi-scale interfusion and conditional speaker modulation (ConSM), which is called MC-SpEx. First of all, we design the weight-share multi-scale fusers (ScaleFusers) for efficiently leveraging multi-scale information as well as ensuring consistency of the model's feature space. Then, to consider different scale information while generating masks, the multi-scale interactive mask generator (ScaleInterMG) is presented. Moreover, we introduce ConSM module to fully exploit speaker embedding in the speech extractor. Experimental results on the Libri2Mix dataset demonstrate the effectiveness of our improvements and the state-of-the-art performance of our proposed MC-SpEx.

Gesper: A Restoration-Enhancement Framework for General Speech Reconstruction

This paper describes a real-time General Speech Reconstruction (Gesper) system submitted to the ICASSP 2023 Speech Signal Improvement (SSI) Challenge. This novel proposed system is a two-stage architecture, in which the speech restoration is performed, and then cascaded by speech enhancement. We propose a complex spectral mapping-based generative adversarial network (CSM-GAN) as the speech restoration module for the first time. For noise suppression and dereverberation, the enhancement module is performed with fullband-wideband parallel processing. On the blind test set of ICASSP 2023 SSI Challenge, the proposed Gesper system, which satisfies the real-time condition, achieves 3.27 P.804 overall mean opinion score (MOS) and 3.35 P.835 overall MOS, ranked 1st in both track 1 and track 2.

TEA-PSE 3.0: Tencent-Ethereal-Audio-Lab Personalized Speech Enhancement System For ICASSP 2023 DNS Challenge

This paper introduces the Unbeatable Team's submission to the ICASSP 2023 Deep Noise Suppression (DNS) Challenge. We expand our previous work, TEA-PSE, to its upgraded version -- TEA-PSE 3.0. Specifically, TEA-PSE 3.0 incorporates a residual LSTM after squeezed temporal convolution network (S-TCN) to enhance sequence modeling capabilities. Additionally, the local-global representation (LGR) structure is introduced to boost speaker information extraction, and multi-STFT resolution loss is used to effectively capture the time-frequency characteristics of the speech signals. Moreover, retraining methods are employed based on the freeze training strategy to fine-tune the system. According to the official results, TEA-PSE 3.0 ranks 1st in both ICASSP 2023 DNS-Challenge track 1 and track 2.

ExARN: self-attending RNN for target speaker extraction

Target speaker extraction is to extract the target speaker, specified by enrollment utterance, in an environment with other competing speakers. Therefore, the task needs to solve two problems, speaker identification and separation, at the same time. In this paper, we combine self-attention and Recurrent Neural Networks (RNN). Further, we exploit various ways to combining different auxiliary information with mixed representations. Experimental results show that our proposed model achieves excellent performance on the task of target speaker extraction.

Speech Enhancement with Intelligent Neural Homomorphic Synthesis

Most neural network speech enhancement models ignore speech production mathematical models by directly mapping Fourier transform spectrums or waveforms. In this work, we propose a neural source filter network for speech enhancement. Specifically, we use homomorphic signal processing and cepstral analysis to obtain noisy speech's excitation and vocal tract. Unlike traditional signal processing, we use an attentive recurrent network (ARN) model predicted ratio mask to replace the liftering separation function. Then two convolutional attentive recurrent network (CARN) networks are used to predict the excitation and vocal tract of clean speech, respectively. The system's output is synthesized from the estimated excitation and vocal. Experiments prove that our proposed method performs better, with SI-SNR improving by 1.363dB compared to FullSubNet.

Local-global speaker representation for target speaker extraction

Target speaker extraction is to extract the target speaker's voice from a mixture of signals according to the given enrollment utterance. The target speaker's enrollment utterance is also called as anchor speech. The effective utilization of anchor speech is crucial for speaker extraction. In this study, we propose a new system to exploit speaker information from anchor speech fully. Unlike models that use only local or global features of the anchor, the proposed method extracts speaker information on global and local levels and feeds the features into a speech separation network. Our approach benefits from the complementary advantages of both global and local features, and the performance of speaker extraction is improved. We verified the feasibility of this local-global representation (LGR) method using multiple speaker extraction models. Systematic experiments were conducted on the open-source dataset Libri-2talker, and the results showed that the proposed method significantly outperformed the baseline models.