Search

In this work, we propose a novel consistency-preserving loss function for recovering the phase information in the context of phase reconstruction (PR) and speech enhancement (SE). Different from conventional techniques that directly estimate the phase using a deep model, our idea is to exploit ad-hoc constraints to directly generate a consistent pair of magnitude and phase. Specifically, the proposed loss forces a set of complex numbers to be a consistent short-time Fourier transform (STFT) representation, i.e., to be the spectrogram of a real signal. Our approach thus avoids the difficulty of estimating the original phase, which is highly unstructured and sensitive to time shift. The influence of our proposed loss is first assessed on a PR task, experimentally demonstrating that our approach is viable. Next, we show its effectiveness on an SE task, using both the VB-DMD and WSJ0-CHiME3 data sets. On VB-DMD, our approach is competitive with conventional solutions. On the challenging WSJ0-CHiME3 set, the proposed framework compares favourably over those techniques that explicitly estimate the phase., Comment: 5 pages, Submitted to ICASSP 2025

This work investigates two strategies for zero-shot non-intrusive speech assessment leveraging large language models. First, we explore the audio analysis capabilities of GPT-4o. Second, we propose GPT-Whisper, which uses Whisper as an audio-to-text module and evaluates the naturalness of text via targeted prompt engineering. We evaluate assessment metrics predicted by GPT-4o and GPT-Whisper examining their correlations with human-based quality and intelligibility assessments, and character error rate (CER) of automatic speech recognition. Experimental results show that GPT-4o alone is not effective for audio analysis; whereas, GPT-Whisper demonstrates higher prediction, showing moderate correlation with speech quality and intelligibility, and high correlation with CER. Compared to supervised non-intrusive neural speech assessment models, namely MOS-SSL and MTI-Net, GPT-Whisper yields a notably higher Spearman's rank correlation with the CER of Whisper. These findings validate GPT-Whisper as a reliable method for accurate zero-shot speech assessment without requiring additional training data (speech data and corresponding assessment scores).

Given recent advances in generative AI technology, a key question is how large language models (LLMs) can enhance acoustic modeling tasks using text decoding results from a frozen, pretrained automatic speech recognition (ASR) model. To explore new capabilities in language modeling for speech processing, we introduce the generative speech transcription error correction (GenSEC) challenge. This challenge comprises three post-ASR language modeling tasks: (i) post-ASR transcription correction, (ii) speaker tagging, and (iii) emotion recognition. These tasks aim to emulate future LLM-based agents handling voice-based interfaces while remaining accessible to a broad audience by utilizing open pretrained language models or agent-based APIs. We also discuss insights from baseline evaluations, as well as lessons learned for designing future evaluations., Comment: IEEE SLT 2024. The initial draft version has been done in December 2023. Post-ASR Text Processing and Understanding Community and LlaMA-7B pre-training correction model: https://huggingface.co/GenSEC-LLM/SLT-Task1-Llama2-7b-HyPo-baseline

Self-supervised representation learning (SSL) has attained SOTA results on several downstream speech tasks, but SSL-based speech enhancement (SE) solutions still lag behind. To address this issue, we exploit three main ideas: (i) Transformer-based masking generation, (ii) consistency-preserving loss, and (iii) perceptual contrast stretching (PCS). In detail, conformer layers, leveraging an attention mechanism, are introduced to effectively model frame-level representations and obtain the Ideal Ratio Mask (IRM) for SE. Moreover, we incorporate consistency in the loss function, which processes the input to account for the inconsistency effects of signal reconstruction from the spectrogram. Finally, PCS is employed to improve the contrast of input and target features according to perceptual importance. Evaluated on the VoiceBank-DEMAND task, the proposed solution outperforms previously SSL-based SE solutions when tested on several objective metrics, attaining a SOTA PESQ score of 3.54.

This work is concerned with devising a robust Parkinson's (PD) disease detector from speech in real-world operating conditions using (i) foundational models, and (ii) speech enhancement (SE) methods. To this end, we first fine-tune several foundational-based models on the standard PC-GITA (s-PC-GITA) clean data. Our results demonstrate superior performance to previously proposed models. Second, we assess the generalization capability of the PD models on the extended PC-GITA (e-PC-GITA) recordings, collected in real-world operative conditions, and observe a severe drop in performance moving from ideal to real-world conditions. Third, we align training and testing conditions applaying off-the-shelf SE techniques on e-PC-GITA, and a significant boost in performance is observed only for the foundational-based models. Finally, combining the two best foundational-based models trained on s-PC-GITA, namely WavLM Base and Hubert Base, yielded top performance on the enhanced e-PC-GITA., Comment: Accepted at INTERSPEECH 2024

Italy exhibits rich linguistic diversity across its territory due to the distinct regional languages spoken in different areas. Recent advances in self-supervised learning provide new opportunities to analyze Italy's linguistic varieties using speech data alone. This includes the potential to leverage representations learned from large amounts of data to better examine nuances between closely related linguistic varieties. In this study, we focus on automatically identifying the geographic region of origin of speech samples drawn from Italy's diverse language varieties. We leverage self-supervised learning models to tackle this task and analyze differences and similarities between Italy's regional languages. In doing so, we also seek to uncover new insights into the relationships among these diverse yet closely related varieties, which may help linguists understand their interconnected evolution and regional development over time and space. To improve the discriminative ability of learned representations, we evaluate several supervised contrastive learning objectives, both as pre-training steps and additional fine-tuning objectives. Experimental evidence shows that pre-trained self-supervised models can effectively identify regions from speech recording. Additionally, incorporating contrastive objectives during fine-tuning improves classification accuracy and yields embeddings that distinctly separate regional varieties, demonstrating the value of combining self-supervised pre-training and contrastive learning for this task., Comment: Accepted to LREC-COLING 2024 - https://aclanthology.org/2024.lrec-main.1317/

We propose a novel language-universal approach to end-to-end automatic spoken keyword recognition (SKR) leveraging upon (i) a self-supervised pre-trained model, and (ii) a set of universal speech attributes (manner and place of articulation). Specifically, Wav2Vec2.0 is used to generate robust speech representations, followed by a linear output layer to produce attribute sequences. A non-trainable pronunciation model then maps sequences of attributes into spoken keywords in a multilingual setting. Experiments on the Multilingual Spoken Words Corpus show comparable performances to character- and phoneme-based SKR in seen languages. The inclusion of domain adversarial training (DAT) improves the proposed framework, outperforming both character- and phoneme-based SKR approaches with 13.73% and 17.22% relative word error rate (WER) reduction in seen languages, and achieves 32.14% and 19.92% WER reduction for unseen languages in zero-shot settings.

This work aims to study a scalable state-space model (SSM), Mamba, for the speech enhancement (SE) task. We exploit a Mamba-based regression model to characterize speech signals and build an SE system upon Mamba, termed SEMamba. We explore the properties of Mamba by integrating it as the core model in both basic and advanced SE systems, along with utilizing signal-level distances as well as metric-oriented loss functions. SEMamba demonstrates promising results and attains a PESQ score of 3.55 on the VoiceBank-DEMAND dataset. When combined with the perceptual contrast stretching technique, the proposed SEMamba yields a new state-of-the-art PESQ score of 3.69.

Limited diversity in standardized benchmarks for evaluating audio representation learning (ARL) methods may hinder systematic comparison of current methods' capabilities. We present ARCH, a comprehensive benchmark for evaluating ARL methods on diverse audio classification domains, covering acoustic events, music, and speech. ARCH comprises 12 datasets, that allow us to thoroughly assess pre-trained SSL models of different sizes. ARCH streamlines benchmarking of ARL techniques through its unified access to a wide range of domains and its ability to readily incorporate new datasets and models. To address the current lack of open-source, pre-trained models for non-speech audio, we also release new pre-trained models that demonstrate strong performance on non-speech datasets. We argue that the presented wide-ranging evaluation provides valuable insights into state-of-the-art ARL methods, and is useful to pinpoint promising research directions.

Recent studies have successfully shown that large language models (LLMs) can be successfully used for generative error correction (GER) on top of the automatic speech recognition (ASR) output. Specifically, an LLM is utilized to carry out a direct mapping from the N-best hypotheses list generated by an ASR system to the predicted output transcription. However, despite its effectiveness, GER introduces extra data uncertainty since the LLM is trained without taking into account acoustic information available in the speech signal. In this work, we aim to overcome such a limitation by infusing acoustic information before generating the predicted transcription through a novel late fusion solution termed Uncertainty-Aware Dynamic Fusion (UADF). UADF is a multimodal fusion approach implemented into an auto-regressive decoding process and works in two stages: (i) It first analyzes and calibrates the token-level LLM decision, and (ii) it then dynamically assimilates the information from the acoustic modality. Experimental evidence collected from various ASR tasks shows that UADF surpasses existing fusion mechanisms in several ways. It yields significant improvements in word error rate (WER) while mitigating data uncertainty issues in LLM and addressing the poor generalization relied with sole modality during fusion. We also demonstrate that UADF seamlessly adapts to audio-visual speech recognition., Comment: Accepted to ICLR 2024, 17 pages. This work will be open sourced under MIT license

In this work, we aim to establish a Bayesian adaptive learning framework by focusing on estimating latent variables in deep neural network (DNN) models. Latent variables indeed encode both transferable distributional information and structural relationships. Thus the distributions of the source latent variables (prior) can be combined with the knowledge learned from the target data (likelihood) to yield the distributions of the target latent variables (posterior) with the goal of addressing acoustic mismatches between training and testing conditions. The prior knowledge transfer is accomplished through Variational Bayes (VB). In addition, we also investigate Maximum a Posteriori (MAP) based Bayesian adaptation. Experimental results on device adaptation in acoustic scene classification show that our proposed approaches can obtain good improvements on target devices, and consistently outperforms other cut-edging algorithms., Comment: ASRU2023 Bayesian Symposium. arXiv admin note: text overlap with arXiv:2110.08598

Code-switching (CS) speech refers to the phenomenon of mixing two or more languages within the same sentence. Despite the recent advances in automatic speech recognition (ASR), CS-ASR is still a challenging task ought to the grammatical structure complexity of the phenomenon and the data scarcity of specific training corpus. In this work, we propose to leverage large language models (LLMs) and lists of hypotheses generated by an ASR to address the CS problem. Specifically, we first employ multiple well-trained ASR models for N-best hypotheses generation, with the aim of increasing the diverse and informative elements in the set of hypotheses. Next, we utilize the LLMs to learn the hypotheses-to-transcription (H2T) mapping by adding a trainable low-rank adapter. Such a generative error correction (GER) method directly predicts the accurate transcription according to its expert linguistic knowledge and N-best hypotheses, resulting in a paradigm shift from the traditional language model rescoring or error correction techniques. Experimental evidence demonstrates that GER significantly enhances CS-ASR accuracy, in terms of reduced mixed error rate (MER). Furthermore, LLMs show remarkable data efficiency for H2T learning, providing a potential solution to the data scarcity problem of CS-ASR in low-resource languages., Comment: Submitted to ICASSP2024

Advancements in deep neural networks have allowed automatic speech recognition (ASR) systems to attain human parity on several publicly available clean speech datasets. However, even state-of-the-art ASR systems experience performance degradation when confronted with adverse conditions, as a well-trained acoustic model is sensitive to variations in the speech domain, e.g., background noise. Intuitively, humans address this issue by relying on their linguistic knowledge: the meaning of ambiguous spoken terms is usually inferred from contextual cues thereby reducing the dependency on the auditory system. Inspired by this observation, we introduce the first open-source benchmark to utilize external large language models (LLMs) for ASR error correction, where N-best decoding hypotheses provide informative elements for true transcription prediction. This approach is a paradigm shift from the traditional language model rescoring strategy that can only select one candidate hypothesis as the output transcription. The proposed benchmark contains a novel dataset, HyPoradise (HP), encompassing more than 334,000 pairs of N-best hypotheses and corresponding accurate transcriptions across prevalent speech domains. Given this dataset, we examine three types of error correction techniques based on LLMs with varying amounts of labeled hypotheses-transcription pairs, which gains a significant word error rate (WER) reduction. Experimental evidence demonstrates the proposed technique achieves a breakthrough by surpassing the upper bound of traditional re-ranking based methods. More surprisingly, LLM with reasonable prompt and its generative capability can even correct those tokens that are missing in N-best list. We make our results publicly accessible for reproducible pipelines with released pre-trained models, thus providing a new evaluation paradigm for ASR error correction with LLMs., Comment: Accepted to NeurIPS 2023, 24 pages. Datasets and Benchmarks Track. Added the first Mandarin and code-switching (zh-cn and en-us) results from the LLM-based generative ASR error correction to Table 8 on Page 21

We propose a first step toward multilingual end-to-end automatic speech recognition (ASR) by integrating knowledge about speech articulators. The key idea is to leverage a rich set of fundamental units that can be defined "universally" across all spoken languages, referred to as speech attributes, namely manner and place of articulation. Specifically, several deterministic attribute-to-phoneme mapping matrices are constructed based on the predefined set of universal attribute inventory, which projects the knowledge-rich articulatory attribute logits, into output phoneme logits. The mapping puts knowledge-based constraints to limit inconsistency with acoustic-phonetic evidence in the integrated prediction. Combined with phoneme recognition, our phone recognizer is able to infer from both attribute and phoneme information. The proposed joint multilingual model is evaluated through phoneme recognition. In multilingual experiments over 6 languages on benchmark datasets LibriSpeech and CommonVoice, we find that our proposed solution outperforms conventional multilingual approaches with a relative improvement of 6.85% on average, and it also demonstrates a much better performance compared to monolingual model. Further analysis conclusively demonstrates that the proposed solution eliminates phoneme predictions that are inconsistent with attributes.

Previous Multimodal Information based Speech Processing (MISP) challenges mainly focused on audio-visual speech recognition (AVSR) with commendable success. However, the most advanced back-end recognition systems often hit performance limits due to the complex acoustic environments. This has prompted a shift in focus towards the Audio-Visual Target Speaker Extraction (AVTSE) task for the MISP 2023 challenge in ICASSP 2024 Signal Processing Grand Challenges. Unlike existing audio-visual speech enhance-ment challenges primarily focused on simulation data, the MISP 2023 challenge uniquely explores how front-end speech processing, combined with visual clues, impacts back-end tasks in real-world scenarios. This pioneering effort aims to set the first benchmark for the AVTSE task, offering fresh insights into enhancing the ac-curacy of back-end speech recognition systems through AVTSE in challenging and real acoustic environments. This paper delivers a thorough overview of the task setting, dataset, and baseline system of the MISP 2023 challenge. It also includes an in-depth analysis of the challenges participants may encounter. The experimental results highlight the demanding nature of this task, and we look forward to the innovative solutions participants will bring forward., Comment: 5 pages, 4 figures

Contemporary real-world online ad auctions differ from canonical models [Edelman et al., 2007; Varian, 2009] in at least four ways: (1) values and click-through rates can depend upon users' search queries, but advertisers can only partially "tune" their bids to specific queries; (2) advertisers do not know the number, identity, and precise value distribution of competing bidders; (3) advertisers only receive partial, aggregated feedback, and (4) payment rules are only partially known to bidders. These features make it virtually impossible to fully characterize equilibrium bidding behavior. This paper shows that, nevertheless, one can still gain useful insight into modern ad auctions by modeling advertisers as agents governed by an adversarial bandit algorithm, independent of auction mechanism intricacies. To demonstrate our approach, we first simulate "soft-floor" auctions [Zeithammer, 2019], a complex, real-world pricing rule for which no complete equilibrium characterization is known. We find that (i) when values and click-through rates are query-dependent, soft floors can improve revenues relative to standard auction formats even if bidder types are drawn from the same distribution; and (ii) with distributional asymmetries that reflect relevant real-world scenario, we find that soft floors yield lower revenues than suitably chosen reserve prices, even restricting attention to a single query. We then demonstrate how to infer advertiser value distributions from observed bids for a variety of pricing rules, and illustrate our approach with aggregate data from an e-commerce website.

We address the video prediction task by putting forth a novel model that combines (i) our recently proposed hierarchical residual vector quantized variational autoencoder (HR-VQVAE), and (ii) a novel spatiotemporal PixelCNN (ST-PixelCNN). We refer to this approach as a sequential hierarchical residual learning vector quantized variational autoencoder (S-HR-VQVAE). By leveraging the intrinsic capabilities of HR-VQVAE at modeling still images with a parsimonious representation, combined with the ST-PixelCNN's ability at handling spatiotemporal information, S-HR-VQVAE can better deal with chief challenges in video prediction. These include learning spatiotemporal information, handling high dimensional data, combating blurry prediction, and implicit modeling of physical characteristics. Extensive experimental results on the KTH Human Action and Moving-MNIST tasks demonstrate that our model compares favorably against top video prediction techniques both in quantitative and qualitative evaluations despite a much smaller model size. Finally, we boost S-HR-VQVAE by proposing a novel training method to jointly estimate the HR-VQVAE and ST-PixelCNN parameters., Comment: 14 pages, 7 figures, 3 tables. Submitted to IEEE Transactions on Pattern Analysis and Machine Intelligence on 2023-07-12

Using kinematic properties of handwriting to support the diagnosis of neurodegenerative disease is a real challenge: non-invasive detection techniques combined with machine learning approaches promise big steps forward in this research field. In literature, the tasks proposed focused on different cognitive skills to elicitate handwriting movements. In particular, the meaning and phonology of words to copy can compromise writing fluency. In this paper, we investigated how word semantics and phonology affect the handwriting of people affected by Alzheimer's disease. To this aim, we used the data from six handwriting tasks, each requiring copying a word belonging to one of the following categories: regular (have a predictable phoneme-grapheme correspondence, e.g., cat), non-regular (have atypical phoneme-grapheme correspondence, e.g., laugh), and non-word (non-meaningful pronounceable letter strings that conform to phoneme-grapheme conversion rules). We analyzed the data using a machine learning approach by implementing four well-known and widely-used classifiers and feature selection. The experimental results showed that the feature selection allowed us to derive a different set of highly distinctive features for each word type. Furthermore, non-regular words needed, on average, more features but achieved excellent classification performance: the best result was obtained on a non-regular, reaching an accuracy close to 90%.

We propose a multi-dimensional structured state space (S4) approach to speech enhancement. To better capture the spectral dependencies across the frequency axis, we focus on modifying the multi-dimensional S4 layer with whitening transformation to build new small-footprint models that also achieve good performance. We explore several S4-based deep architectures in time (T) and time-frequency (TF) domains. The 2-D S4 layer can be considered a particular convolutional layer with an infinite receptive field although it utilizes fewer parameters than a conventional convolutional layer. Evaluated on the VoiceBank-DEMAND data set, when compared with the conventional U-net model based on convolutional layers, the proposed TF-domain S4-based model is 78.6% smaller in size, yet it still achieves competitive results with a PESQ score of 3.15 with data augmentation. By increasing the model size, we can even reach a PESQ score of 3.18., Comment: Accepted to Interspeech 2023. Code will be released at https://github.com/Kuray107/S4ND-U-Net_speech_enhancement

In this work, we devise a parameter-efficient solution to bring differential privacy (DP) guarantees into adaptation of a cross-lingual speech classifier. We investigate a new frozen pre-trained adaptation framework for DP-preserving speech modeling without full model fine-tuning. First, we introduce a noisy teacher-student ensemble into a conventional adaptation scheme leveraging a frozen pre-trained acoustic model and attain superior performance than DP-based stochastic gradient descent (DPSGD). Next, we insert residual adapters (RA) between layers of the frozen pre-trained acoustic model. The RAs reduce training cost and time significantly with a negligible performance drop. Evaluated on the open-access Multilingual Spoken Words (MLSW) dataset, our solution reduces the number of trainable parameters by 97.5% using the RAs with only a 4% performance drop with respect to fine-tuning the cross-lingual speech classifier while preserving DP guarantees., Comment: Accepted to Interspeech 2023. Code will be available at: https://github.com/Chun-wei-Ho/Private-Speech-Adapter. The authors would like to express their gratitude to Prof. Chin-Hui Lee from Georgia Tech for providing helpful insights and suggestions

The Multi-modal Information based Speech Processing (MISP) challenge aims to extend the application of signal processing technology in specific scenarios by promoting the research into wake-up words, speaker diarization, speech recognition, and other technologies. The MISP2022 challenge has two tracks: 1) audio-visual speaker diarization (AVSD), aiming to solve ``who spoken when'' using both audio and visual data; 2) a novel audio-visual diarization and recognition (AVDR) task that focuses on addressing ``who spoken what when'' with audio-visual speaker diarization results. Both tracks focus on the Chinese language, and use far-field audio and video in real home-tv scenarios: 2-6 people communicating each other with TV noise in the background. This paper introduces the dataset, track settings, and baselines of the MISP2022 challenge. Our analyses of experiments and examples indicate the good performance of AVDR baseline system, and the potential difficulties in this challenge due to, e.g., the far-field video quality, the presence of TV noise in the background, and the indistinguishable speakers., Comment: 5 pages, 4 figures, to be published in ICASSP2023

We propose a quantum kernel learning (QKL) framework to address the inherent data sparsity issues often encountered in training large-scare acoustic models in low-resource scenarios. We project acoustic features based on classical-to-quantum feature encoding. Different from existing quantum convolution techniques, we utilize QKL with features in the quantum space to design kernel-based classifiers. Experimental results on challenging spoken command recognition tasks for a few low-resource languages, such as Arabic, Georgian, Chuvash, and Lithuanian, show that the proposed QKL-based hybrid approach attains good improvements over existing classical and quantum solutions., Comment: Submitted to ICASSP 2023

This study addresses the speech enhancement (SE) task within the causal inference paradigm by modeling the noise presence as an intervention. Based on the potential outcome framework, the proposed causal inference-based speech enhancement (CISE) separates clean and noisy frames in an intervened noisy speech using a noise detector and assigns both sets of frames to two mask-based enhancement modules (EMs) to perform noise-conditional SE. Specifically, we use the presence of noise as guidance for EM selection during training, and the noise detector selects the enhancement module according to the prediction of the presence of noise for each frame. Moreover, we derived a SE-specific average treatment effect to quantify the causal effect adequately. Experimental evidence demonstrates that CISE outperforms a non-causal mask-based SE approach in the studied settings and has better performance and efficiency than more complex SE models., Comment: Submitted to ICASSP 2023

We propose an ensemble learning framework with Poisson sub-sampling to effectively train a collection of teacher models to issue some differential privacy (DP) guarantee for training data. Through boosting under DP, a student model derived from the training data suffers little model degradation from the models trained with no privacy protection. Our proposed solution leverages upon two mechanisms, namely: (i) a privacy budget amplification via Poisson sub-sampling to train a target prediction model that requires less noise to achieve a same level of privacy budget, and (ii) a combination of the sub-sampling technique and an ensemble teacher-student learning framework that introduces DP-preserving noise at the output of the teacher models and transfers DP-preserving properties via noisy labels. Privacy-preserving student models are then trained with the noisy labels to learn the knowledge with DP-protection from the teacher model ensemble. Experimental evidences on spoken command recognition and continuous speech recognition of Mandarin speech show that our proposed framework greatly outperforms existing DP-preserving algorithms in both speech processing tasks., Comment: Accepted to ISCA, ISCSLP 2022, Singapore. 5 Pages

Differential privacy (DP) is one data protection avenue to safeguard user information used for training deep models by imposing noisy distortion on privacy data. Such a noise perturbation often results in a severe performance degradation in automatic speech recognition (ASR) in order to meet a privacy budget $\varepsilon$. Private aggregation of teacher ensemble (PATE) utilizes ensemble probabilities to improve ASR accuracy when dealing with the noise effects controlled by small values of $\varepsilon$. We extend PATE learning to work with dynamic patterns, namely speech utterances, and perform a first experimental demonstration that it prevents acoustic data leakage in ASR training. We evaluate three end-to-end deep models, including LAS, hybrid CTC/attention, and RNN transducer, on the open-source LibriSpeech and TIMIT corpora. PATE learning-enhanced ASR models outperform the benchmark DP-SGD mechanisms, especially under strict DP budgets, giving relative word error rate reductions between 26.2% and 27.5% for an RNN transducer model evaluated with LibriSpeech. We also introduce a DP-preserving ASR solution for pretraining on public speech corpora., Comment: 5 pages. Accepted to IEEE SLT 2022. A first version draft was finished in Aug 2021