Recent progress in speech separation has been largely driven by advances in deep neural networks, yet their high computational and memory requirements hinder deployment on resource-constrained devices. A significant inefficiency in conventional systems arises from using static network architectures that maintain constant computational complexity across all input segments, regardless of their characteristics. This approach is sub-optimal for simpler segments that do not require intensive processing, such as silence or non-overlapping speech. To address this limitation, we propose a dynamic slimmable network (DSN) for speech separation that adaptively adjusts its computational complexity based on the input signal. The DSN combines a slimmable network, which can operate at different network widths, with a lightweight gating module that dynamically determines the required width by analyzing the local input characteristics. To balance performance and efficiency, we introduce a signal-dependent complexity loss that penalizes unnecessary computation based on segmental reconstruction error. Experiments on clean and noisy two-speaker mixtures from the WSJ0-2mix and WHAM! datasets show that the DSN achieves a better performance-efficiency trade-off than individually trained static networks of different sizes.
We extend the frameworks of Serialized Output Training (SOT) to address practical needs of both streaming and offline automatic speech recognition (ASR) applications. Our approach focuses on balancing latency and accuracy, catering to real-time captioning and summarization requirements. We propose several key improvements: (1) Leveraging Continuous Speech Separation (CSS) single-channel front-end with end-to-end (E2E) systems for highly overlapping scenarios, challenging the conventional wisdom of E2E versus cascaded setups. The CSS framework improves the accuracy of the ASR system by separating overlapped speech from multiple speakers. (2) Implementing dual models -- Conformer Transducer for streaming and Sequence-to-Sequence for offline -- or alternatively, a two-pass model based on cascaded encoders. (3) Exploring segment-based SOT (segSOT) which is better suited for offline scenarios while also enhancing readability of multi-talker transcriptions.
Target Speech Extraction (TSE) aims to isolate a target speaker's voice from a mixture of multiple speakers by leveraging speaker-specific cues, typically provided as auxiliary audio (a.k.a. cue audio). Although recent advancements in TSE have primarily employed discriminative models that offer high perceptual quality, these models often introduce unwanted artifacts, reduce naturalness, and are sensitive to discrepancies between training and testing environments. On the other hand, generative models for TSE lag in perceptual quality and intelligibility. To address these challenges, we present SoloSpeech, a novel cascaded generative pipeline that integrates compression, extraction, reconstruction, and correction processes. SoloSpeech features a speaker-embedding-free target extractor that utilizes conditional information from the cue audio's latent space, aligning it with the mixture audio's latent space to prevent mismatches. Evaluated on the widely-used Libri2Mix dataset, SoloSpeech achieves the new state-of-the-art intelligibility and quality in target speech extraction and speech separation tasks while demonstrating exceptional generalization on out-of-domain data and real-world scenarios.
This paper addresses the problem of single-channel speech separation, where the number of speakers is unknown, and each speaker may speak multiple utterances. We propose a speech separation model that simultaneously performs separation, dynamically estimates the number of speakers, and detects individual speaker activities by integrating an attractor module. The proposed system outperforms existing methods by introducing an attractor-based architecture that effectively combines local and global temporal modeling for multi-utterance scenarios. To evaluate the method in reverberant and noisy conditions, a multi-speaker multi-utterance dataset was synthesized by combining Librispeech speech signals with WHAM! noise signals. The results demonstrate that the proposed system accurately estimates the number of sources. The system effectively detects source activities and separates the corresponding utterances into correct outputs in both known and unknown source count scenarios.
This paper aims to achieve single-channel target speech extraction (TSE) in enclosures utilizing distance clues and room information. Recent works have verified the feasibility of distance clues for the TSE task, which can imply the sound source's direct-to-reverberation ratio (DRR) and thus can be utilized for speech separation and TSE systems. However, such distance clue is significantly influenced by the room's acoustic characteristics, such as dimension and reverberation time, making it challenging for TSE systems that rely solely on distance clues to generalize across a variety of different rooms. To solve this, we suggest providing room environmental information (room dimensions and reverberation time) for distance-based TSE for better generalization capabilities. Especially, we propose a distance and environment-based TSE model in the time-frequency (TF) domain with learnable distance and room embedding. Results on both simulated and real collected datasets demonstrate its feasibility. Demonstration materials are available at https://runwushi.github.io/distance-room-demo-page/.
Although deep learning has substantially advanced speech separation in recent years, most existing studies continue to prioritize separation quality while overlooking computational efficiency, an essential factor for low-latency speech processing in real-time applications. In this paper, we propose SepPrune, the first structured pruning framework specifically designed to compress deep speech separation models and reduce their computational cost. SepPrune begins by analyzing the computational structure of a given model to identify layers with the highest computational burden. It then introduces a differentiable masking strategy to enable gradient-driven channel selection. Based on the learned masks, SepPrune prunes redundant channels and fine-tunes the remaining parameters to recover performance. Extensive experiments demonstrate that this learnable pruning paradigm yields substantial advantages for channel pruning in speech separation models, outperforming existing methods. Notably, a model pruned with SepPrune can recover 85% of the performance of a pre-trained model (trained over hundreds of epochs) with only one epoch of fine-tuning, and achieves convergence 36$\times$ faster than training from scratch. Code is available at https://github.com/itsnotacie/SepPrune.
Existing causal speech separation models often underperform compared to non-causal models due to difficulties in retaining historical information. To address this, we propose the Time-Frequency Attention Cache Memory (TFACM) model, which effectively captures spatio-temporal relationships through an attention mechanism and cache memory (CM) for historical information storage. In TFACM, an LSTM layer captures frequency-relative positions, while causal modeling is applied to the time dimension using local and global representations. The CM module stores past information, and the causal attention refinement (CAR) module further enhances time-based feature representations for finer granularity. Experimental results showed that TFACM achieveed comparable performance to the SOTA TF-GridNet-Causal model, with significantly lower complexity and fewer trainable parameters. For more details, visit the project page: https://cslikai.cn/TFACM/.
Recent advancements in deep learning have significantly impacted the field of speech signal processing, particularly in the analysis and manipulation of complex spectrograms. This survey provides a comprehensive overview of the state-of-the-art techniques leveraging deep neural networks for processing complex spectrograms, which encapsulate both magnitude and phase information. We begin by introducing complex spectrograms and their associated features for various speech processing tasks. Next, we explore the key components and architectures of complex-valued neural networks, which are specifically designed to handle complex-valued data and have been applied for complex spectrogram processing. We then discuss various training strategies and loss functions tailored for training neural networks to process and model complex spectrograms. The survey further examines key applications, including phase retrieval, speech enhancement, and speech separation, where deep learning has achieved significant progress by leveraging complex spectrograms or their derived feature representations. Additionally, we examine the intersection of complex spectrograms with generative models. This survey aims to serve as a valuable resource for researchers and practitioners in the field of speech signal processing and complex-valued neural networks.
Blind Speech Separation (BSS) aims to separate multiple speech sources from audio mixtures recorded by a microphone array. The problem is challenging because it is a blind inverse problem, i.e., the microphone array geometry, the room impulse response (RIR), and the speech sources, are all unknown. We propose ArrayDPS to solve the BSS problem in an unsupervised, array-agnostic, and generative manner. The core idea builds on diffusion posterior sampling (DPS), but unlike DPS where the likelihood is tractable, ArrayDPS must approximate the likelihood by formulating a separate optimization problem. The solution to the optimization approximates room acoustics and the relative transfer functions between microphones. These approximations, along with the diffusion priors, iterate through the ArrayDPS sampling process and ultimately yield separated voice sources. We only need a simple single-speaker speech diffusion model as a prior along with the mixtures recorded at the microphones; no microphone array information is necessary. Evaluation results show that ArrayDPS outperforms all baseline unsupervised methods while being comparable to supervised methods in terms of SDR. Audio demos are provided at: https://arraydps.github.io/ArrayDPSDemo/.
This paper presents an efficient visual speech encoder for lip reading. While most recent lip reading studies have been based on the ResNet architecture and have achieved significant success, they are not sufficiently suitable for efficiently capturing lip reading features due to high computational complexity in modeling spatio-temporal information. Additionally, using a complex visual model not only increases the complexity of lip reading models but also induces delays in the overall network for multi-modal studies (e.g., audio-visual speech recognition, speech enhancement, and speech separation). To overcome the limitations of Convolutional Neural Network (CNN)-based models, we apply the hierarchical structure and window self-attention of the Swin Transformer to lip reading. We configure a new lightweight scale of the Swin Transformer suitable for processing lip reading data and present the SwinLip visual speech encoder, which efficiently reduces computational load by integrating modified Convolution-augmented Transformer (Conformer) temporal embeddings with conventional spatial embeddings in the hierarchical structure. Through extensive experiments, we have validated that our SwinLip successfully improves the performance and inference speed of the lip reading network when applied to various backbones for word and sentence recognition, reducing computational load. In particular, our SwinLip demonstrated robust performance in both English LRW and Mandarin LRW-1000 datasets and achieved state-of-the-art performance on the Mandarin LRW-1000 dataset with less computation compared to the existing state-of-the-art model.
While contemporary speech separation technologies adeptly process lengthy mixed audio waveforms, they are frequently challenged by the intricacies of real-world environments, including noisy and reverberant settings, which can result in artifacts or distortions in the separated speech. To overcome these limitations, we introduce SepALM, a pioneering approach that employs audio language models (ALMs) to rectify and re-synthesize speech within the text domain following preliminary separation. SepALM comprises four core components: a separator, a corrector, a synthesizer, and an aligner. By integrating an ALM-based end-to-end error correction mechanism, we mitigate the risk of error accumulation and circumvent the optimization hurdles typically encountered in conventional methods that amalgamate automatic speech recognition (ASR) with large language models (LLMs). Additionally, we have developed Chain-of-Thought (CoT) prompting and knowledge distillation techniques to facilitate the reasoning and training processes of the ALM. Our experiments substantiate that SepALM not only elevates the precision of speech separation but also markedly bolsters adaptability in novel acoustic environments.
Signal separation in the passive underwater acoustic domain has heavily relied on deep learning techniques to isolate ship radiated noise. However, the separation networks commonly used in this domain stem from speech separation applications and may not fully consider the unique aspects of underwater acoustics beforehand, such as the influence of different propagation media, signal frequencies and modulation characteristics. This oversight highlights the need for tailored approaches that account for the specific characteristics of underwater sound propagation. This study introduces a novel temporal network designed to separate ship radiated noise by employing a dual-path model and a feature decoupling approach. The mixed signals' features are transformed into a space where they exhibit greater independence, with each dimension's significance decoupled. Subsequently, a fusion of local and global attention mechanisms is employed in the separation layer. Extensive comparisons showcase the effectiveness of this method when compared to other prevalent network models, as evidenced by its performance in the ShipsEar and DeepShip datasets.
Speech separation (SS) seeks to disentangle a multi-talker speech mixture into single-talker speech streams. Although SS can be generally achieved using offline methods, such a processing paradigm is not suitable for real-time streaming applications. Causal separation models, which rely only on past and present information, offer a promising solution for real-time streaming. However, these models typically suffer from notable performance degradation due to the absence of future context. In this paper, we introduce a novel frontend that is designed to mitigate the mismatch between training and run-time inference by implicitly incorporating future information into causal models through predictive patterns. The pretrained frontend employs a transformer decoder network with a causal convolutional encoder as the backbone and is pretrained in a self-supervised manner with two innovative pretext tasks: autoregressive hybrid prediction and contextual knowledge distillation. These tasks enable the model to capture predictive patterns directly from mixtures in a self-supervised manner. The pretrained frontend subsequently serves as a feature extractor to generate high-quality predictive patterns. Comprehensive evaluations on synthetic and real-world datasets validated the effectiveness of the proposed pretrained frontend.
Voice data is increasingly being used in modern digital communications, yet there is still a lack of comprehensive tools for automated voice analysis and characterization. To this end, we developed the VANPY (Voice Analysis in Python) framework for automated pre-processing, feature extraction, and classification of voice data. The VANPY is an open-source end-to-end comprehensive framework that was developed for the purpose of speaker characterization from voice data. The framework is designed with extensibility in mind, allowing for easy integration of new components and adaptation to various voice analysis applications. It currently incorporates over fifteen voice analysis components - including music/speech separation, voice activity detection, speaker embedding, vocal feature extraction, and various classification models. Four of the VANPY's components were developed in-house and integrated into the framework to extend its speaker characterization capabilities: gender classification, emotion classification, age regression, and height regression. The models demonstrate robust performance across various datasets, although not surpassing state-of-the-art performance. As a proof of concept, we demonstrate the framework's ability to extract speaker characteristics on a use-case challenge of analyzing character voices from the movie "Pulp Fiction." The results illustrate the framework's capability to extract multiple speaker characteristics, including gender, age, height, emotion type, and emotion intensity measured across three dimensions: arousal, dominance, and valence.
Generative models have attracted considerable attention for speech separation tasks, and among these, diffusion-based methods are being explored. Despite the notable success of diffusion techniques in generation tasks, their adaptation to speech separation has encountered challenges, notably slow convergence and suboptimal separation outcomes. To address these issues and enhance the efficacy of diffusion-based speech separation, we introduce EDSep, a novel single-channel method grounded in score matching via stochastic differential equation (SDE). This method enhances generative modeling for speech source separation by optimizing training and sampling efficiency. Specifically, a novel denoiser function is proposed to approximate data distributions, which obtains ideal denoiser outputs. Additionally, a stochastic sampler is carefully designed to resolve the reverse SDE during the sampling process, gradually separating speech from mixtures. Extensive experiments on databases such as WSJ0-2mix, LRS2-2mix, and VoxCeleb2-2mix demonstrate our proposed method's superior performance over existing diffusion and discriminative models, validating its efficacy.
Speech separation approaches for single-channel, dry speech mixtures have significantly improved. However, real-world spatial and reverberant acoustic environments remain challenging, limiting the effectiveness of these approaches for assistive hearing devices like cochlear implants (CIs). To address this, we quantify the impact of real-world acoustic scenes on speech separation and explore how spatial cues can enhance separation quality efficiently. We analyze performance based on implicit spatial cues (inherent in the acoustic input and learned by the model) and explicit spatial cues (manually calculated spatial features added as auxiliary inputs). Our findings show that spatial cues (both implicit and explicit) improve separation for mixtures with spatially separated and nearby talkers. Furthermore, spatial cues enhance separation when spectral cues are ambiguous, such as when voices are similar. Explicit spatial cues are particularly beneficial when implicit spatial cues are weak. For instance, single CI microphone recordings provide weaker implicit spatial cues than bilateral CIs, but even single CIs benefit from explicit cues. These results emphasize the importance of training models on real-world data to improve generalizability in everyday listening scenarios. Additionally, our statistical analyses offer insights into how data properties influence model performance, supporting the development of efficient speech separation approaches for CIs and other assistive devices in real-world settings.
Target Speech Extraction (TSE) traditionally relies on explicit clues about the speaker's identity like enrollment audio, face images, or videos, which may not always be available. In this paper, we propose a text-guided TSE model StyleTSE that uses natural language descriptions of speaking style in addition to the audio clue to extract the desired speech from a given mixture. Our model integrates a speech separation network adapted from SepFormer with a bi-modality clue network that flexibly processes both audio and text clues. To train and evaluate our model, we introduce a new dataset TextrolMix with speech mixtures and natural language descriptions. Experimental results demonstrate that our method effectively separates speech based not only on who is speaking, but also on how they are speaking, enhancing TSE in scenarios where traditional audio clues are absent. Demos are at: https://mingyue66.github.io/TextrolMix/demo/
The goal of this paper is speech separation and enhancement in multi-speaker and noisy environments using a combination of different modalities. Previous works have shown good performance when conditioning on temporal or static visual evidence such as synchronised lip movements or face identity. In this paper, we present a unified framework for multi-modal speech separation and enhancement based on synchronous or asynchronous cues. To that end we make the following contributions: (i) we design a modern Transformer-based architecture tailored to fuse different modalities to solve the speech separation task in the raw waveform domain; (ii) we propose conditioning on the textual content of a sentence alone or in combination with visual information; (iii) we demonstrate the robustness of our model to audio-visual synchronisation offsets; and, (iv) we obtain state-of-the-art performance on the well-established benchmark datasets LRS2 and LRS3.
The topic of speech separation involves separating mixed speech with multiple overlapping speakers into several streams, with each stream containing speech from only one speaker. Many highly effective models have emerged and proliferated rapidly over time. However, the size and computational load of these models have also increased accordingly. This is a disaster for the community, as researchers need more time and computational resources to reproduce and compare existing models. In this paper, we propose U-mamba-net: a lightweight Mamba-based U-style model for speech separation in complex environments. Mamba is a state space sequence model that incorporates feature selection capabilities. U-style network is a fully convolutional neural network whose symmetric contracting and expansive paths are able to learn multi-resolution features. In our work, Mamba serves as a feature filter, alternating with U-Net. We test the proposed model on Libri2mix. The results show that U-Mamba-Net achieves improved performance with quite low computational cost.
Neural audio codecs have revolutionized audio processing by enabling speech tasks to be performed on highly compressed representations. Recent work has shown that speech separation can be achieved within these compressed domains, offering faster training and reduced inference costs. However, current approaches still rely on waveform-based loss functions, necessitating unnecessary decoding steps during training. We propose a novel embedding loss for neural audio codec-based speech separation that operates directly on compressed audio representations, eliminating the need for decoding during training. To validate our approach, we conduct comprehensive evaluations using both objective metrics and perceptual assessment techniques, including intrusive and non-intrusive methods. Our results demonstrate that embedding loss can be used to train codec-based speech separation models with a 2x improvement in training speed and computational cost while achieving better DNSMOS and STOI performance on the WSJ0-2mix dataset across 3 different pre-trained codecs.