Adaptability of ASR Models on Low-Resource Language: A Comparative Study of Whisper and Wav2Vec-BERT on Bangla

In recent years, neural models trained on large multilingual text and speech datasets have shown great potential for supporting low-resource languages. This study investigates the performances of two state-of-the-art Automatic Speech Recognition (ASR) models, OpenAI's Whisper (Small & Large-V2) and Facebook's Wav2Vec-BERT on Bangla, a low-resource language. We have conducted experiments using two publicly available datasets: Mozilla Common Voice-17 and OpenSLR to evaluate model performances. Through systematic fine-tuning and hyperparameter optimization, including learning rate, epochs, and model checkpoint selection, we have compared the models based on Word Error Rate (WER), Character Error Rate (CER), Training Time, and Computational Efficiency. The Wav2Vec-BERT model outperformed Whisper across all key evaluation metrics, demonstrated superior performance while requiring fewer computational resources, and offered valuable insights to develop robust speech recognition systems in low-resource linguistic settings.

Perceptual Ratings Predict Speech Inversion Articulatory Kinematics in Childhood Speech Sound Disorders

Purpose: This study evaluated whether articulatory kinematics, inferred by Articulatory Phonology speech inversion neural networks, aligned with perceptual ratings of /r/ and /s/ in the speech of children with speech sound disorders. Methods: Articulatory Phonology vocal tract variables were inferred for 5,961 utterances from 118 children and 3 adults, aged 2.25-45 years. Perceptual ratings were standardized using the novel 5-point PERCEPT Rating Scale and training protocol. Two research questions examined if the articulatory patterns of inferred vocal tract variables aligned with the perceptual error category for the phones investigated (e.g., tongue tip is more anterior in dentalized /s/ productions than in correct /s/). A third research question examined if gradient PERCEPT Rating Scale scores predicted articulatory proximity to correct productions. Results: Estimated marginal means from linear mixed models supported 17 of 18 /r/ hypotheses, involving tongue tip and tongue body constrictions. For /s/, estimated marginal means from a second linear mixed model supported 7 of 15 hypotheses, particularly those related to the tongue tip. A third linear mixed model revealed that PERCEPT Rating Scale scores significantly predicted articulatory proximity of errored phones to correct productions. Conclusion: Inferred vocal tract variables differentiated category and magnitude of articulatory errors for /r/, and to a lesser extent for /s/, aligning with perceptual judgments. These findings support the clinical interpretability of speech inversion vocal tract variables and the PERCEPT Rating Scale in quantifying articulatory proximity to the target sound, particularly for /r/.

A Dataset for Automatic Assessment of TTS Quality in Spanish

This work addresses the development of a database for the automatic assessment of text-to-speech (TTS) systems in Spanish, aiming to improve the accuracy of naturalness prediction models. The dataset consists of 4,326 audio samples from 52 different TTS systems and human voices and is, up to our knowledge, the first of its kind in Spanish. To label the audios, a subjective test was designed based on the ITU-T Rec. P.807 standard and completed by 92 participants. Furthermore, the utility of the collected dataset was validated by training automatic naturalness prediction systems. We explored two approaches: fine-tuning an existing model originally trained for English, and training small downstream networks on top of frozen self-supervised speech models. Our models achieve a mean absolute error of 0.8 on a five-point MOS scale. Further analysis demonstrates the quality and diversity of the developed dataset, and its potential to advance TTS research in Spanish.

First Steps Towards Voice Anonymization for Code-Switching Speech

The goal of voice anonymization is to modify an audio such that the true identity of its speaker is hidden. Research on this task is typically limited to the same English read speech datasets, thus the efficacy of current methods for other types of speech data remains unknown. In this paper, we present the first investigation of voice anonymization for the multilingual phenomenon of code-switching speech. We prepare two corpora for this task and propose adaptations to a multilingual anonymization model to make it applicable for code-switching speech. By testing the anonymization performance of this and two language-independent methods on the datasets, we find that only the multilingual system performs well in terms of privacy and utility preservation. Furthermore, we observe challenges in performing utility evaluations on this data because of its spontaneous character and the limited code-switching support by the multilingual speech recognition model.

A Review on Sound Source Localization in Robotics: Focusing on Deep Learning Methods

Sound source localization (SSL) adds a spatial dimension to auditory perception, allowing a system to pinpoint the origin of speech, machinery noise, warning tones, or other acoustic events, capabilities that facilitate robot navigation, human-machine dialogue, and condition monitoring. While existing surveys provide valuable historical context, they typically address general audio applications and do not fully account for robotic constraints or the latest advancements in deep learning. This review addresses these gaps by offering a robotics-focused synthesis, emphasizing recent progress in deep learning methodologies. We start by reviewing classical methods such as Time Difference of Arrival (TDOA), beamforming, Steered-Response Power (SRP), and subspace analysis. Subsequently, we delve into modern machine learning (ML) and deep learning (DL) approaches, discussing traditional ML and neural networks (NNs), convolutional neural networks (CNNs), convolutional recurrent neural networks (CRNNs), and emerging attention-based architectures. The data and training strategy that are the two cornerstones of DL-based SSL are explored. Studies are further categorized by robot types and application domains to facilitate researchers in identifying relevant work for their specific contexts. Finally, we highlight the current challenges in SSL works in general, regarding environmental robustness, sound source multiplicity, and specific implementation constraints in robotics, as well as data and learning strategies in DL-based SSL. Also, we sketch promising directions to offer an actionable roadmap toward robust, adaptable, efficient, and explainable DL-based SSL for next-generation robots.

MambAttention: Mamba with Multi-Head Attention for Generalizable Single-Channel Speech Enhancement

With the advent of new sequence models like Mamba and xLSTM, several studies have shown that these models match or outperform state-of-the-art models in single-channel speech enhancement, automatic speech recognition, and self-supervised audio representation learning. However, prior research has demonstrated that sequence models like LSTM and Mamba tend to overfit to the training set. To address this issue, previous works have shown that adding self-attention to LSTMs substantially improves generalization performance for single-channel speech enhancement. Nevertheless, neither the concept of hybrid Mamba and time-frequency attention models nor their generalization performance have been explored for speech enhancement. In this paper, we propose a novel hybrid architecture, MambAttention, which combines Mamba and shared time- and frequency-multi-head attention modules for generalizable single-channel speech enhancement. To train our model, we introduce VoiceBank+Demand Extended (VB-DemandEx), a dataset inspired by VoiceBank+Demand but with more challenging noise types and lower signal-to-noise ratios. Trained on VB-DemandEx, our proposed MambAttention model significantly outperforms existing state-of-the-art LSTM-, xLSTM-, Mamba-, and Conformer-based systems of similar complexity across all reported metrics on two out-of-domain datasets: DNS 2020 and EARS-WHAM_v2, while matching their performance on the in-domain dataset VB-DemandEx. Ablation studies highlight the role of weight sharing between the time- and frequency-multi-head attention modules for generalization performance. Finally, we explore integrating the shared time- and frequency-multi-head attention modules with LSTM and xLSTM, which yields a notable performance improvement on the out-of-domain datasets. However, our MambAttention model remains superior on both out-of-domain datasets across all reported evaluation metrics.

GGTalker: Talking Head Systhesis with Generalizable Gaussian Priors and Identity-Specific Adaptation

Creating high-quality, generalizable speech-driven 3D talking heads remains a persistent challenge. Previous methods achieve satisfactory results for fixed viewpoints and small-scale audio variations, but they struggle with large head rotations and out-of-distribution (OOD) audio. Moreover, they are constrained by the need for time-consuming, identity-specific training. We believe the core issue lies in the lack of sufficient 3D priors, which limits the extrapolation capabilities of synthesized talking heads. To address this, we propose GGTalker, which synthesizes talking heads through a combination of generalizable priors and identity-specific adaptation. We introduce a two-stage Prior-Adaptation training strategy to learn Gaussian head priors and adapt to individual characteristics. We train Audio-Expression and Expression-Visual priors to capture the universal patterns of lip movements and the general distribution of head textures. During the Customized Adaptation, individual speaking styles and texture details are precisely modeled. Additionally, we introduce a color MLP to generate fine-grained, motion-aligned textures and a Body Inpainter to blend rendered results with the background, producing indistinguishable, photorealistic video frames. Comprehensive experiments show that GGTalker achieves state-of-the-art performance in rendering quality, 3D consistency, lip-sync accuracy, and training efficiency.

Aligning Spoken Dialogue Models from User Interactions

We propose a novel preference alignment framework for improving spoken dialogue models on real-time conversations from user interactions. Current preference learning methods primarily focus on text-based language models, and are not directly suited to the complexities of real-time speech interactions, with richer dynamics (e.g. interruption, interjection) and no explicit segmentation between speaker turns.We create a large-scale dataset of more than 150,000 preference pairs from raw multi-turn speech conversations, annotated with AI feedback, to cover preferences over both linguistic content and temporal context variations. We leverage offline alignment methods to finetune a full-duplex autoregressive speech-to-speech model. Extensive experiments demonstrate that feedback on generic conversations can be consistently effective in improving spoken dialogue models to produce more factual, safer and more contextually aligned interactions. We deploy the finetuned model and conduct holistic human evaluations to assess the impact beyond single-turn conversations. Our findings shed light on the importance of a well-calibrated balance among various dynamics, crucial for natural real-time speech dialogue systems.

Deception Detection in Dyadic Exchanges Using Multimodal Machine Learning: A Study on a Swedish Cohort

This study investigates the efficacy of using multimodal machine learning techniques to detect deception in dyadic interactions, focusing on the integration of data from both the deceiver and the deceived. We compare early and late fusion approaches, utilizing audio and video data - specifically, Action Units and gaze information - across all possible combinations of modalities and participants. Our dataset, newly collected from Swedish native speakers engaged in truth or lie scenarios on emotionally relevant topics, serves as the basis for our analysis. The results demonstrate that incorporating both speech and facial information yields superior performance compared to single-modality approaches. Moreover, including data from both participants significantly enhances deception detection accuracy, with the best performance (71%) achieved using a late fusion strategy applied to both modalities and participants. These findings align with psychological theories suggesting differential control of facial and vocal expressions during initial interactions. As the first study of its kind on a Scandinavian cohort, this research lays the groundwork for future investigations into dyadic interactions, particularly within psychotherapy settings.

Hybrid Deep Learning and Signal Processing for Arabic Dialect Recognition in Low-Resource Settings

Arabic dialect recognition presents a significant challenge in speech technology due to the linguistic diversity of Arabic and the scarcity of large annotated datasets, particularly for underrepresented dialects. This research investigates hybrid modeling strategies that integrate classical signal processing techniques with deep learning architectures to address this problem in low-resource scenarios. Two hybrid models were developed and evaluated: (1) Mel-Frequency Cepstral Coefficients (MFCC) combined with a Convolutional Neural Network (CNN), and (2) Discrete Wavelet Transform (DWT) features combined with a Recurrent Neural Network (RNN). The models were trained on a dialect-filtered subset of the Common Voice Arabic dataset, with dialect labels assigned based on speaker metadata. Experimental results demonstrate that the MFCC + CNN architecture achieved superior performance, with an accuracy of 91.2% and strong precision, recall, and F1-scores, significantly outperforming the Wavelet + RNN configuration, which achieved an accuracy of 66.5%. These findings highlight the effectiveness of leveraging spectral features with convolutional models for Arabic dialect recognition, especially when working with limited labeled data. The study also identifies limitations related to dataset size, potential regional overlaps in labeling, and model optimization, providing a roadmap for future research. Recommendations for further improvement include the adoption of larger annotated corpora, integration of self-supervised learning techniques, and exploration of advanced neural architectures such as Transformers. Overall, this research establishes a strong baseline for future developments in Arabic dialect recognition within resource-constrained environments.