期刊文献+
任意字段
题名或关键词
题名
关键词
摘要
作者
第一作者
机构
刊名
分类号
参考文献
作者简介
基金资助
栏目信息
共找到17,264篇文章
< 1 2 … 250 >
每页显示 20 50 100
已选0条
引用分析
参考文献
引证文献
 统计分析
检索结果
已选文献
显示方式：
A Prosody-Guided Multi-Stream Framework for Universal Detection of AI-Synthesized Speech across Codec and Vocoder Domains 认领 引用
1
作者 Akmalbek Abdusalomov Mukhriddin Mukhiddinov +6 位作者 Fakhriddin Abdirazakov Alpamis Kutlimuratov Nodira Alimova Ilyos Kalandarov Ayhan Istanbullu Rashid Nasimov Young-Im Cho 《Computers, Materials & Continua》 SCIE EI 2026年第7期2247-2267,共21页
Recent advancements in AI-synthesized speech have resulted in highly realistic deepfake audio,posing severe threats to authentication systems and digital media trust.Existing detection models struggle to generalize ac... Recent advancements in AI-synthesized speech have resulted in highly realistic deepfake audio,posing severe threats to authentication systems and digital media trust.Existing detection models struggle to generalize across diverse synthesis methods,especially those involving neural codec-based Audio Language Models(ALMs).In this work,we propose UniTector++,a novel prosody-aware,multi-stream detection architecture that generalizes across vocoder-and codec-based synthesis.UniTector++incorporates three complementary streams—Whisper-based semantic embeddings,high-level prosodic features,and codec artifact representations—fused through a Multi-Domain Adaptive Graph Attention Fusion(MAGAF)module.Furthermore,an Emotion-Consistency Verification Module(ECVM)reinforces alignment between speech style and prosodic content,and a Universal Adversarial Robustness(UAR)head improves resistance against adversarial attacks.Evaluated on three benchmark datasets—ASVspoof2021,PolyFake,and Codecfake—UniTector++achieves state-of-the-art performance with average Equal Error Rate(EER)of 0.57%under unseen synthesis scenarios,outperforming competitive baselines by a relative margin of 28%.Our results demonstrate the model’s superior generalization,interpretability,and robustness,offering a significant advancement in universal deepfake speech detection. 展开更多
关键词 Deepfake speech detection prosody analysis neural codec artifacts whisper model multi-stream fusion emotion-consistency verification AI-synthesized speech spoofing detection
暂未订购 下载PDF
Simultaneous Speech Extraction for Multiple Target Speakers Under Meeting Scenarios 认领 引用
2
作者 ZENG Bang SUO Hongbin +1 位作者 WAN Yulong LI Ming 《Journal of Shanghai Jiaotong university(Science)》 EI 2026年第2期258-264,共7页
The common target speech separation directly estimates the target source,ignoring the interrelationship between different speakers at each frame.We propose a multiple-target speech separation(MTSS)model to simultaneou... The common target speech separation directly estimates the target source,ignoring the interrelationship between different speakers at each frame.We propose a multiple-target speech separation(MTSS)model to simultaneously extract each speaker’s voice from the mixed speech rather than just optimally estimating the target source.Moreover,we propose a speaker diarization(SD)aware MTSS system(SD-MTSS).By exploiting the target speaker voice activity detection(TSVAD)and the estimated mask,our SD-MTSS model can extract the speech signal of each speaker concurrently in a conversational recording without additional enrollment audio in advance.Experimental results show that our MTSS model achieves improvements of 1.38 dB signal-to-distortion ratio(SDR),1.34 dB scale-invariant signal-to-distortion ratio(SISDR),and 0.13 perceptual evaluation of speech quality(PESQ)over the baseline on the WSJ0-2mix-extr dataset,separately.The SD-MTSS system makes a 19.2%relative speaker dependent character error rate reduction on the AliMeeting dataset. 展开更多
关键词 target speech separation interrelationship speaker diarization(SD) target speaker voice activity detection multiple-target speech separation(MTSS)model
暂未订购 下载PDF
Fast speech style adaptation with adjustable prosody and variable duration 认领 引用
3
作者 Zhiqiang Hua Lingyun Yu +2 位作者 Chuanbin Liu Dengdi Sun Hongtao Xie 《Journal of University of Science and Technology of China》 CAS CSCD 北大核心 2026年第1期57-68,I0002,共12页
For achieving personalized speech synthesis,it is indispensable to synthesize speech with diverse prosody for any given text.This task presents two key challenges:first,existing methods struggle to simultaneously extr... For achieving personalized speech synthesis,it is indispensable to synthesize speech with diverse prosody for any given text.This task presents two key challenges:first,existing methods struggle to simultaneously extract local prosody information and control phoneme duration,while overlooking the impact of duration on prosody;second,current speaker adaptation approaches suffer from slow learning speed or poor generalization to unseen speakers outside the training set.To address the aforementioned issues,this paper introduces a novel framework.Our method innovatively introduces the text-to-speech alignment mechanism into prosody modeling,using the aligned text-to-duration to segment speech and obtain local prosodic information,and simultaneously training the two components simplifies the workflow.After obtaining the prosodic information,we use it as a condition to guide the model to learn the corresponding phoneme durations under different types of prosody.We combine this style control work with adapter fine-tuning to quickly synthesize speech with the speaker’s style using small amounts of data from unseen speakers in the training set.Experimental results show that our approach is effective in adjusting prosody and variable duration as well as fast style adapter,and the subjective evaluations of the prosodic modulation model considering duration exhibits a significant improvement. 展开更多
关键词 personalized speech synthesis adjustable prosody variable duration speech style adaption
暂未订购 下载PDF
A Hybrid CNN-BiLSTM Framework for Speech Emotion Recognition with TimeGAN-Augmented Data and Contrastive Learning 认领 引用
4
作者 Rashid Jahangir Muhammad Asif Nauman +1 位作者 Oumaima Saidani Faisal Ramzan 《Computers, Materials & Continua》 SCIE EI 2026年第9期775-794,共20页
Speech Emotion Recognition(SER)is a critical component of affective computing with broad applications in human–computer interaction,mental health monitoring,and intelligent multimedia systems.However,SER remains chal... Speech Emotion Recognition(SER)is a critical component of affective computing with broad applications in human–computer interaction,mental health monitoring,and intelligent multimedia systems.However,SER remains challenging due to the emotional ambiguity,lack of labeled data,class imbalance,and speaker variability.This study presents an effective SER framework that integrates contrastive representation learning,optimized spectrogram-based data augmentation,and selective synthetic data generation by using TimeGAN to enhance emotion classification performance.Contrastive learning enables the model to better discriminate acoustically similar emotions while Optuna automatically tunes augmentation strategies such as noise injection,time shifting,and time-frequency masking.Unlike existing approaches that apply synthetic generation uniformly across all classes,the proposed method targets only confusing or under-represented emotion classes to preserve the inter-class separability.A CNN-BiLSTM architecture is used to extract spectral and temporal information of the speech.The framework is evaluated with benchmark SER datasets—EMO-DB and RAVDESS—under speaker independent protocols.Experimental results demonstrate improved accuracy,robustness,and generalization under limited and imbalanced data conditions,supported by confusion matrices,UMAP,and t-SNE visualizations. 展开更多
关键词 Speech emotion recognition data augmentation optuna TimeGAN synthetic data contrastive learning
暂未订购 下载PDF
Research Progress and Trends of Intelligent Speech in Pathological Healthcare 认领 引用
5
作者 GAO Yingming WU Yangqing +3 位作者 YANG Fei ZHOU Yingying LI Ya WU Mengyue 《Journal of Shanghai Jiaotong university(Science)》 EI 2026年第3期671-692,I0023,共22页
Intelligent speech technology,rooted in the ancient medical practice of“auscultation and interrogation”,is emerging as a transformative tool in modern pathological healthcare.By analyzing acoustic biomarkers within ... Intelligent speech technology,rooted in the ancient medical practice of“auscultation and interrogation”,is emerging as a transformative tool in modern pathological healthcare.By analyzing acoustic biomarkers within speech,voice,cough,breath sounds,and heart sounds,it offers a non-invasive,cost-effective avenue for early screening,auxiliary diagnosis,monitoring,and rehabilitation assessment across a wide spectrum of conditions,including mental disorders,neurodegenerative diseases,respiratory illnesses,cardiovascular diseases,and laryngeal or vocal tract pathologies.This study comprehensively reviews the research progress and prevailing trends in this interdisciplinary field.It begins by elucidating the physiological basis of pathological acoustics,including neural dysregulation,airway and pulmonary abnormalities,hemodynamic disturbances,and laryngeal or vocal tract dysfunction,and then discusses its integration with AI-driven diagnostics.The core of the review systematically details advances in two pillars:data resource construction(encompassing datasets for various diseases and standardization efforts)and methodological innovation(tracking the paradigm shift from feature-based machine learning to deep learning,self-supervised models,and multimodal large language models).Furthermore,it explores the development of intelligent speech-driven intervention systems for mental health.The analysis identifies key dynamic trends:the evolution from single-modality to multimodal analysis,the shift from strong to weak/self-supervised learning,the transition from controlled lab settings to naturalistic scenarios,the growing priority of model interpretability,and the move towards multi-disease coexistence modeling.Despite promising clinical potential,significant challenges persist,including data scarcity,algorithmic robustness,and clinical integration bottlenecks.The study concludes by outlining critical future directions:fostering federated learning and multi-center validation,enhancing explainable AI fused with medical knowledge,improving cross-device and cross-environment robustness through hardware-software co-design,refining human-AI collaborative diagnostic paradigms,and establishing comprehensive standardization and regulatory frameworks.Overcoming these hurdles through concerted interdisciplinary efforts is essential to realize a full-cycle intelligent health ecosystem,advancing precision medicine and equitable healthcare delivery. 展开更多
关键词 pathological acoustics intelligent speech analysis bioacoustic biomarker cardiopulmonary sound multimodal learning clinical decision support
暂未订购 下载PDF
RSG-Conformer:ReLU-Based Sparse and Grouped Conformer for Audio-Visual Speech Recognition 认领 引用
6
作者 Yewei Xiao Xin Du Wei Zeng 《Computers, Materials & Continua》 SCIE EI 2026年第3期1325-1348,共24页
Audio-visual speech recognition(AVSR),which integrates audio and visual modalities to improve recognition performance and robustness in noisy or adverse acoustic conditions,has attracted significant research interest.... Audio-visual speech recognition(AVSR),which integrates audio and visual modalities to improve recognition performance and robustness in noisy or adverse acoustic conditions,has attracted significant research interest.However,Conformer-based architectures remain computational expensive due to the quadratic increase in the spatial and temporal complexity of their softmax-based attention mechanisms with sequence length.In addition,Conformerbased architectures may not provide sufficient flexibility for modeling local dependencies at different granularities.To mitigate these limitations,this study introduces a novel AVSR framework based on a ReLU-based Sparse and Grouped Conformer(RSG-Conformer)architecture.Specifically,we propose a Global-enhanced Sparse Attention(GSA)module incorporating an efficient context restoration block to recover lost contextual cues.Concurrently,a Grouped-scale Convolution(GSC)module replaces the standard Conformer convolution module,providing adaptive local modeling across varying temporal resolutions.Furthermore,we integrate a Refined Intermediate Contextual CTC(RIC-CTC)supervision strategy.This approach applies progressively increasing loss weights combined with convolution-based context aggregation,thereby further relaxing the constraint of conditional independence inherent in standard CTC frameworks.Evaluations on the LRS2 and LRS3 benchmark validate the efficacy of our approach,with word error rates(WERs)reduced to 1.8%and 1.5%,respectively.These results further demonstrate and validate its state-of-the-art performance in AVSR tasks. 展开更多
关键词 Audio-visual speech recognition conformer CTC sparse attention
暂未订购 下载PDF
Exploring Generation of Pronunciation Lexicon for Low-Resource Language Automatic Speech Recognition Based on Generic Phone Recognizer 认领 引用
7
作者 LI Jinpeng CHEN Xie ZHANG Weiqiang 《Journal of Shanghai Jiaotong university(Science)》 EI 2026年第2期265-272,共8页
The lexicon is an essential component in the hybrid automatic speech recognition(ASR)system.However,a high-quality lexicon requires significant efforts from the linguistic experts and is difficult to obtain,especially... The lexicon is an essential component in the hybrid automatic speech recognition(ASR)system.However,a high-quality lexicon requires significant efforts from the linguistic experts and is difficult to obtain,especially for low-resource languages.This paper addresses the problem of using a well-trained universal phone recognizer,obtained through the training of multilingual speech data and pronunciation lexicons,to generate pronunciation lexicons for low-resource languages driven by speech data.We propose a simple pipeline that utilizes this approach to generate pronunciation lexicons and apply them into ASR systems.The steps to generate the lexicon are simple and generic:applying the International Phonetic Alphabet(IPA)phone recognizer on the speech,then aligning it with the reference word sequence,followed by filtering to obtain a series of AUTO-subwords,using them to generate the AUTO-subword lexicon and the AUTO-IPA lexicon.We used the pronunciation lexicon generated for the hybrid system and for fine-tuning the pre-trained model.According to the experiment results,we are able to construct the lexicon without resourcing to linguistic experts.Furthermore,the generated lexicon is able to outperform grapheme-based lexicon and is comparable to expert lexicon. 展开更多
关键词 International Phonetic Alphabet(IPA) lexicon learning phone recognition low-resource speech recognition
暂未订购 下载PDF
Improvement of Emotion Detection by Fusing Speech and Image Based on CNN with Temporal Models 认领 引用
8
作者 Shing-Tai Pan Yi-Zhen Huang Zhi-Qing Chen 《Computers, Materials & Continua》 SCIE EI 2026年第9期1677-1697,共21页
This paper proposes a multimodal fusion framework that integrates speech and visual features to enhance the accuracy of emotion recognition.The principal contribution lies in extending the visual component from single... This paper proposes a multimodal fusion framework that integrates speech and visual features to enhance the accuracy of emotion recognition.The principal contribution lies in extending the visual component from single-image to multi-image emotion recognition.Specifically,the proposed framework employs an InceptionV3 Convolutional Neural Network(CNN)-based architecture to extract features from multiple facial images representing the speaker’s expressions throughout an utterance.These features are concatenated into a single vector and subsequently processed by Long Short-Term Memory(LSTM)or Hidden Markov Model(HMM)for temporal modeling.For the speech modality,Mel-Frequency Cepstral Coefficients(MFCC)or filter bank features are extracted from processed audio signals and fed into a hybrid CNN-time-series model.The two modalities are then integrated through model-level and decision-level fusion strategies.Since recognition accuracy tends to degrade as the number of utterances and speakers increases,the Ryerson Audio-Visual Database of Emotional Speech and Song(RAVDESS),which contains a moderate number of sentences and speakers,is adopted in this study.Experimental results demonstrate that the proposed multi-image approach improves recognition accuracy from 91%to 96%compared with the single-image baseline,and that the multimodal fusion framework consistently outperforms its single-modal counterpart. 展开更多
关键词 Speech emotion recognition consecutive facial image emotion recognition convolutional neural network(CNN) long short-term memory(LSTM) hidden markov model(HMM) support vector machine(SVM)
暂未订购 下载PDF
A Comparative Study of Audio-Language Models for Speech EmotionRecognition in Spanish 认领 引用
9
作者 Jorge Gómez-Navalón Ronghao Pan +2 位作者 Tomas Bernal-Beltrán José Antonio García-Díaz Rafael Valencia-García 《Computer Modeling in Engineering & Sciences》 SCIE EI 2026年第7期1096-1123,共28页
Traditionally,speech emotion recognition has relied on supervised models that require task-specific training and annotated data.However,the recent emergence of audio-language models introduces a more flexible paradigm... Traditionally,speech emotion recognition has relied on supervised models that require task-specific training and annotated data.However,the recent emergence of audio-language models introduces a more flexible paradigm that enables multimodal reasoning through speech and natural language interaction.Nevertheless,their effectiveness for emotion recognition remains unclear.In this study,we evaluate audio-language models for speech emotion classification using the Spanish MEACorpus dataset and compare three approaches:prompt-based inference,embedding-based classification with lightweight classifiers,and instruction-tuned models with parameter-efficient fine-tuning plus a hybrid architecture based on class-specific confidence-driven routing.Our results show that the hybrid approach achieves the highest overall performance,reaching an 83.55%macro F1-score and an 84.37%weighted F1-score.Instruction tuning remains highly competitive,obtaining an 83.32%macro F1-score,which confirms the importance of supervised task adaptation for aligning ALMs with speech emotion recognition.We also include Whisper as a pretrained acoustic baseline to contextualize ALM-based representations against strong speech foundation models.Furthermore,the hybrid approach outperforms standalone embedding-based classification across all evaluated models,showing that class-specific confidence-driven routing can improve the use of embedding-based predictions.Although the best hybrid ALM configuration achieves competitive performance,it remains below the specialized MEACorpus baseline of 87.74%macro F1-score,indicating that general-purpose ALMs do not yet surpass highly specialized acoustic models for Spanish speech emotion recognition. 展开更多
关键词 Speech emotion recognition audio-language models multimodal learning instruction tuning prompting embedding-based classification
暂未订购 下载PDF
Wav2vec-AD:Acoustic Unit Discovery Module-Integrated,Self-Supervised Contrastive Pre-training Approach for Speech Recognition 认领 引用
10
作者 NURMEMET Yolwas SUN Lixu +2 位作者 LI Xin LIU Qichao WANG Zhixiang 《Journal of Shanghai Jiaotong university(Science)》 EI 2026年第2期289-297,共9页
An effective speech recognition model necessitates an ample supply of labeled data for supervised training.However,this proposition poses a monumental challenge for low-resource languages in terms of constructing a sp... An effective speech recognition model necessitates an ample supply of labeled data for supervised training.However,this proposition poses a monumental challenge for low-resource languages in terms of constructing a speech recognition system with high precision.In this paper,we propose a novel pre-training strategy for contrastive learning by fusing the acoustic unit discovery module with Wav2vec 2.0,herein referred to asWav2vec-AD.This strategy,for the first time in speech contrastive learning,enables controlled negative sample selection via the acoustic unit discovery module,thereby augmenting the model’s representational learning capability.Furthermore,we conduct a thorough analysis regarding the selection of negative samples in different situations to enhance the speech representation learned by the model,optimizing its efficacy in downstream tasks.In the low-resource case,compared to the baseline Wav2vec 2.0,Wav2vec-AD achieves absolute word error rate(WER)improvements of 1.55%and 1.46%respectively on the development-clean and test-clean subsets of LibriSpeech.Moreover,absolute WER improvements of 0.63%and 4.21%were realized in Arabic and Turkish language datasets,respectively. 展开更多
关键词 self-supervised learning automatic speech recognition contrastive learning low-resource
暂未订购 下载PDF
EC-BERT:A BERT Language Model with Error Correction for Mandarin Chinese Speech Recognition 认领 引用
11
作者 XIAO Sujie HAO Ruipeng +2 位作者 CHENG Gaofeng XU Xiaoyan LI Ta 《Journal of Shanghai Jiaotong university(Science)》 EI 2026年第2期282-288,共7页
The attention-based encoder-decoder end-to-end model has achieved promising performance in automatic speech recognition(ASR).However,in practical applications,substitution errors commonly occur in ASR systems,particul... The attention-based encoder-decoder end-to-end model has achieved promising performance in automatic speech recognition(ASR).However,in practical applications,substitution errors commonly occur in ASR systems,particularly for characters with the same or similar pronunciation.According to statistics,homophones cause at least 50%character errors.Therefore,our study focuses on addressing the issue of substitution errors with the same or similar pronunciation.In this study,we propose a BERT language model with error correction(EC-BERT)for the ASR system.We design a two-stage training schedule involving pre-training with a large amount of pseudo-paired data followed by fine-tuning with a small real-paired data to mitigate the inconsistency of the original pre-trained BERT model with our task.Unlike other error correction models,we do not need an error detection network or mask mechanism but directly use the BERT model to learn and correct the error locations.The experimental results show that our proposed method is effective and achieves a relative reduction of 19.2%in character error rate compared with the connectionist temporal classification(CTC)greedy search result and 12.8%compared with the CTC-WFST result on the AISHELL-1 test set.We also prove that our proposed EC-BERT model can achieve comparable results to other error correction models with a shorter runtime and can easily be integrated into the practical ASR system. 展开更多
关键词 automatic speech recognition(ASR) end-to-end BERT error correction
暂未订购 下载PDF
NestLipGNN:A Hierarchical Graph Neural Network Framework with Nested Multi-Granularity Learning for Robust Visual Speech Recognition 认领 引用
12
作者 Vinh Truong Hoang Nghia Dinh +5 位作者 Luu Quang Phuong Kiet Tran-Trung Ha Duong Thi Hong Bay Nguyen Van Hau Nguyen Trung Thien Ho Huong 《Computers, Materials & Continua》 SCIE EI 2026年第7期1287-1310,共24页
Visual speech recognition(VSR)aims to infer spoken content from visual observations of articulatory movements.Despite significant progress,it remains a challenging task in computer vision and speech processing.Its dif... Visual speech recognition(VSR)aims to infer spoken content from visual observations of articulatory movements.Despite significant progress,it remains a challenging task in computer vision and speech processing.Its difficulty arises from pronounced speaker-to-speaker variability,the presence of homophenes(phonemes that are visually indistinguishable),changes in illumination,and the intrinsically high-dimensional nature of spatiotemporal lip dynamics.In this work,we propose NestLipGNN,a graph-based framework that integrates Graph Neural Networks(GNNs)with a nested multi-granularity learning strategy for visual speech recognition.We construct dynamic lip graphs from facial landmarks to model both spatial relationships between lip regions and their temporal motion during speech articulation.The proposed nested learning architecture supports hierarchical feature extraction across several levels of linguistic abstraction,spanning phoneme-level articulatory units,viseme-level visual speech categories,and word-level semantic representations.We further introduce a Temporal Graph Attention mechanism(T-GAT)that adaptively reweights the importance of distinct lip regions over time.We also introduce a graph-based contrastive learning objective to improve the discrimination of visually similar speech patterns,directly confronting the challenge of homophene resolution.Experiments on the LRW,LRS2,LRS3,and GRID datasets show that NestLipGNN improves recognition accuracy compared with existing methods,obtaining 92.3%word-level accuracy on LRW and delivering a 2.1%absolute performance gain over prior methods.Comprehensive ablation analyses confirm the contribution of each architectural component. 展开更多
关键词 Visual speech recognition graph neural networks nested optimization hierarchical representation learning spatiotemporal modeling contrastive learning lip reading
暂未订购 下载PDF
Broca's area,responsible for speech production,is regulated by lung function 认领 引用
13
作者 Siyu Cao Wenwen Zhuang +2 位作者 Yuqian Hu Shijun Qiu Li-Hai Tan 《Neural Regeneration Research》 SCIE CAS CSCD 2026年第9期4352-4357,共6页
For more than 150 years,Broca's area—specifically,the pars opercularis and pars triangularis in the left inferior frontal gyrus—has been recognized as crucial for human speech production.However,it remains unkno... For more than 150 years,Broca's area—specifically,the pars opercularis and pars triangularis in the left inferior frontal gyrus—has been recognized as crucial for human speech production.However,it remains unknown why this region is recruited for speaking.Speech production involves not only conceptualization and motor planning but also respiration to provide the necessary airflow for creating sounds.Thus,the role of Broca's area in speech may be shaped by the functionality and related brain regions of the lungs.To test this hypothesis,we recruited patients with chronic obstructive pulmonary disease and asked them to read words aloud while their brains were scanned using functional magnetic resonance imaging,with quantitative magnetic resonance imaging acquired separately.The chronic obstructive pulmonary disease patients exhibited altered cortical responses in the left inferior prefrontal cortex and other regions during speech tasks,and also had abnormal activation in cortical sites associated with breathing.In addition,using quantitative magnetic resonance imaging to generate longitudinal relaxation time(T1)maps as an index of brain microstructural changes,including dendritic maturation and myelination,we observed significantly longer longitudinal relaxation times in Broca's area in the chronic obstructive pulmonary disease group than in the control group,suggesting reduced myelination and impaired microstructural integrity.Crucially,our data indicated that more severe dyspnea was associated with less well-developed microstructure in Broca's area and weaker activation of this region.The present study indicates for the first time that the lungs may function to shape Broca's area as the speaking center,providing novel evidence that pulmonary dysfunction can influence both the functional and structural properties of cortical language regions.These findings highlight the mechanistic role of the lung-brain axis in speech production and suggest potential targets for interventions to enhance speech performance. 展开更多
关键词 Broca’s area chronic obstructive pulmonary disease dyspnea functional magnetic resonance imaging lung-brain axis speech production lung functions microstructural changes quantitative magnetic resonance imaging
暂未订购 下载PDF
Detection of Maliciously Disseminated Hate Speech in Spanish Using Fine-Tuning and In-Context Learning Techniques with Large Language Models 认领 引用
14
作者 Tomás Bernal-Beltrán RonghaoPan +3 位作者 JoséAntonio García-Díaz María del Pilar Salas-Zárate Mario Andrés Paredes-Valverde Rafael Valencia-García 《Computers, Materials & Continua》 SCIE EI 2026年第4期353-390,共38页
The malicious dissemination of hate speech via compromised accounts,automated bot networks and malware-driven social media campaigns has become a growing cybersecurity concern.Automatically detecting such content in S... The malicious dissemination of hate speech via compromised accounts,automated bot networks and malware-driven social media campaigns has become a growing cybersecurity concern.Automatically detecting such content in Spanish is challenging due to linguistic complexity and the scarcity of annotated resources.In this paper,we compare two predominant AI-based approaches for the forensic detection of malicious hate speech:(1)finetuning encoder-only models that have been trained in Spanish and(2)In-Context Learning techniques(Zero-and Few-Shot Learning)with large-scale language models.Our approach goes beyond binary classification,proposing a comprehensive,multidimensional evaluation that labels each text by:(1)type of speech,(2)recipient,(3)level of intensity(ordinal)and(4)targeted group(multi-label).Performance is evaluated using an annotated Spanish corpus,standard metrics such as precision,recall and F1-score and stability-oriented metrics to evaluate the stability of the transition from zero-shot to few-shot prompting(Zero-to-Few Shot Retention and Zero-to-Few Shot Gain)are applied.The results indicate that fine-tuned encoder-only models(notably MarIA and BETO variants)consistently deliver the strongest and most reliable performance:in our experiments their macro F1-scores lie roughly in the range of approximately 46%–66%depending on the task.Zero-shot approaches are much less stable and typically yield substantially lower performance(observed F1-scores range approximately 0%–39%),often producing invalid outputs in practice.Few-shot prompting(e.g.,Qwen 38B,Mistral 7B)generally improves stability and recall relative to pure zero-shot,bringing F1-scores into a moderate range of approximately 20%–51%but still falling short of fully fine-tuned models.These findings highlight the importance of supervised adaptation and discuss the potential of both paradigms as components in AI-powered cybersecurity and malware forensics systems designed to identify and mitigate coordinated online hate campaigns. 展开更多
关键词 Hate speech detection malicious communication campaigns AI-driven cybersecurity social media analytics large language models prompt-tuning fine-tuning in-context learning natural language processing
暂未订购 下载PDF
Speech Emotion Recognition Based on the Adaptive Acoustic Enhancement and Refined Attention Mechanism 认领 引用
15
作者 Jun Li Chunyan Liang +1 位作者 Zhiguo Liu Fengpei Ge 《Computers, Materials & Continua》 SCIE EI 2026年第3期2015-2039,共25页
To enhance speech emotion recognition capability,this study constructs a speech emotion recognition model integrating the adaptive acoustic mixup(AAM)and improved coordinate and shuffle attention(ICASA)methods.The AAM... To enhance speech emotion recognition capability,this study constructs a speech emotion recognition model integrating the adaptive acoustic mixup(AAM)and improved coordinate and shuffle attention(ICASA)methods.The AAM method optimizes data augmentation by combining a sample selection strategy and dynamic interpolation coefficients,thus enabling information fusion of speech data with different emotions at the acoustic level.The ICASA method enhances feature extraction capability through dynamic fusion of the improved coordinate attention(ICA)and shuffle attention(SA)techniques.The ICA technique reduces computational overhead by employing depth-separable convolution and an h-swish activation function and captures long-range dependencies of multi-scale time-frequency features using the attention weights.The SA technique promotes feature interaction through channel shuffling,which helps the model learn richer and more discriminative emotional features.Experimental results demonstrate that,compared to the baseline model,the proposed model improves the weighted accuracy by 5.42%and 4.54%,and the unweighted accuracy by 3.37%and 3.85%on the IEMOCAP and RAVDESS datasets,respectively.These improvements were confirmed to be statistically significant by independent samples t-tests,further supporting the practical reliability and applicability of the proposed model in real-world emotion-aware speech systems. 展开更多
关键词 Speech emotion recognition adaptive acoustic mixup enhancement improved coordinate attention shuffle attention attention mechanism deep learning
暂未订购 下载PDF
TENG-Based Self-Powered Silent Speech Recognition Interface:from Assistive Communication to Immersive AR/VR Interaction 认领 引用
16
作者 Shuai Lin Yanmin Guo +4 位作者 Xiangyao Zeng Xiongtu Zhou Yongai Zhang Chengda Li Chaoxing Wu 《Nano-Micro Letters》 SCIE EI CAS CSCD 2026年第5期31-44,共14页
Lip language provides a silent,intuitive,and efficient mode of communication,offering a promising solution for individuals with speech impairments.Its articulation relies on complex movements of the jaw and the muscle... Lip language provides a silent,intuitive,and efficient mode of communication,offering a promising solution for individuals with speech impairments.Its articulation relies on complex movements of the jaw and the muscles surrounding it.However,the accurate and real-time acquisition and decoding of these movements into reliable silent speech signals remains a significant challenge.In this work,we propose a real-time silent speech recognition system,which integrates a triboelectric nanogenerator-based flexible pressure sensor(FPS)with a deep learning framework.The FPS employs a porous pyramid-structured silicone film as the negative triboelectric layer,enabling highly sensitive pressure detection in the low-force regime(1 V N-1 for 0-10 N and 4.6 V N-1 for 10-24 N).This allows it to precisely capture jaw movements during speech and convert them into electrical signals.To decode the signals,we proposed a convolutional neural networklong short-term memory(CNN-LSTM)hybrid network,combining CNN and LSTM model to extract both local spatial features and temporal dynamics.The model achieved 95.83%classification accuracy in 30 categories of daily words.Furthermore,the decoded silent speech signals can be directly translated into executable commands for contactless and precise control of the smartphone.The system can also be connected to AR glasses,offering a novel human-machine interaction approach with promising potential in AR/VR applications. 展开更多
关键词 Flexible pressure sensor Silent speech recognition Triboelectric nanogenerator Deep learning AR/VR interaction
暂未订购 下载PDF
A Synthetic Speech Detection Model Combining Local-Global Dependency 认领 引用
17
作者 Jiahui Song Yuepeng Zhang Wenhao Yuan 《Computers, Materials & Continua》 SCIE EI 2026年第1期1312-1326,共15页
Synthetic speech detection is an essential task in the field of voice security,aimed at identifying deceptive voice attacks generated by text-to-speech(TTS)systems or voice conversion(VC)systems.In this paper,we propo... Synthetic speech detection is an essential task in the field of voice security,aimed at identifying deceptive voice attacks generated by text-to-speech(TTS)systems or voice conversion(VC)systems.In this paper,we propose a synthetic speech detection model called TFTransformer,which integrates both local and global features to enhance detection capabilities by effectively modeling local and global dependencies.Structurally,the model is divided into two main components:a front-end and a back-end.The front-end of the model uses a combination of SincLayer and two-dimensional(2D)convolution to extract high-level feature maps(HFM)containing local dependency of the input speech signals.The back-end uses time-frequency Transformer module to process these feature maps and further capture global dependency.Furthermore,we propose TFTransformer-SE,which incorporates a channel attention mechanism within the 2D convolutional blocks.This enhancement aims to more effectively capture local dependencies,thereby improving the model’s performance.The experiments were conducted on the ASVspoof 2021 LA dataset,and the results showed that the model achieved an equal error rate(EER)of 3.37%without data augmentation.Additionally,we evaluated the model using the ASVspoof 2019 LA dataset,achieving an EER of 0.84%,also without data augmentation.This demonstrates that combining local and global dependencies in the time-frequency domain can significantly improve detection accuracy. 展开更多
关键词 Synthetic speech detection transformer local-global time-frequency domain
暂未订购 下载PDF
Multi-Frame Cross-Channel Attention and Speaker Diarization Based Speaker-Attributed Automatic Speech Recognition System for Multi-Channel Multi-Party Meeting Transcription 认领 引用
18
作者 XU Luzhen YAN Haoyin +5 位作者 HE Maokui GUO Zixian ZHOU Yeping LIU Peiqi ZHANG Jie DAI Lirong 《Journal of Shanghai Jiaotong university(Science)》 EI 2026年第2期298-304,共7页
This paper describes a speaker-attributed automatic speech recognition(SA-ASR)system submitted to the multi-channel multi-party meeting transcription challenge,which aims to address the“who spoke what”problem.We ali... This paper describes a speaker-attributed automatic speech recognition(SA-ASR)system submitted to the multi-channel multi-party meeting transcription challenge,which aims to address the“who spoke what”problem.We align the serialized output training-based multi-speaker ASR hypotheses and speaker diarization(SD)results to obtain speaker-attributed transcriptions.We use a pre-trained multi-frame cross-channel attention(MFCCA)model as the ASR module.We build a cascade system which includes a pre-trained speaker overlapaware neural diarization and target-speaker voice activity detection model as the SD module.Decoding and alignment strategies are further used to improve the SA-ASR performance.Our proposed system outperforms the baseline with a relative improvement of 40.3%in terms of concatenated minimum-permutation character error rate on the AliMeeting dataset,which ranks top-3 on the fixed sub-track. 展开更多
关键词 multi-channel multi-party meeting transcription speaker-attributed automatic speech recognition(SA-ASR) serialized output training speaker diarization concatenated minimum-permutation character error rate
暂未订购 下载PDF
Semantic-Sentiment Fusion with Deep Learning:A Novel Framework for Hate Speech Detection 认领 引用
19
作者 Choongwon Kang Haein Lee Jang Hyun Kim 《Computers, Materials & Continua》 SCIE EI 2026年第7期1343-1368,共26页
With the rapid growth of social media and frequent anonymous interactions,hate speech has become widespread.As users express diverse opinions in digital spaces,the need for effective detection remains crucial.To addre... With the rapid growth of social media and frequent anonymous interactions,hate speech has become widespread.As users express diverse opinions in digital spaces,the need for effective detection remains crucial.To address this,we propose a framework applicable to diverse hate speech types,combining sentence-level semantic representation vectors from the pre-trained Bidirectional Encoder Representations from Transformers(BERT)with sentiment score vectors from the Linguistic Inquiry and Word Count(LIWC)dictionary and the Valence Aware Dictionary for sEntiment Reasoning(VADER).This semantic-sentiment fusion integrates three deep learning models—Convolutional Neural Network(CNN),Long Short-Term Memory(LSTM),and Deep Neural Network(DNN)to enhance detection effectiveness.To verify generalizability,we used four datasets:two binary hate speech detection tasks,two multi-class tasks,and validation on another domain dataset.Results show that the proposed framework achieved the best performance,with accuracy up to 91.34%.This approach provides valuable direction for future research. 展开更多
关键词 Hate speech detection natural language processing(NLP) deep learning sentiment analysis
暂未订购 下载PDF
Unraveling Predictive Mechanism in Speech Perception and Production:Insights from EEG Analyses of Brain Network Dynamics 认领 引用
20
作者 ZHAO Bin DANG Jianwu LI Aijun 《Journal of Shanghai Jiaotong university(Science)》 EI 2026年第2期273-281,共9页
How neural networks coordinate to support speech perception and speech production represents a forefront research topic in both contemporary neuroscience and artificial intelligence.Despite the successful incorporatio... How neural networks coordinate to support speech perception and speech production represents a forefront research topic in both contemporary neuroscience and artificial intelligence.Despite the successful incorporation of hierarchical and predictive attributes from biological neural networks(BNNs)into artificial counterparts,substantial disparities persist,particularly in terms of real-time feedback and nonlinear regulation.To gain a more profound understanding of how BNNs manifest these attributes,the present study employed electroencephalography(EEG)techniques to examine the spatiotemporal brain network dynamics involved in listening and oral reading of identical sentences.These two tasks engage distinct sensorimotor modalities while sharing high-level semantic and syntactic representations.According to a hierarchical feedforward model,the low-level auditory and visual inputs would be progressively transformed towards abstract representations of the sentence meaning,leading to a convergence of brain network patterns in higher cognitive regions.However,our findings challenged this viewpoint by revealing an early resemblance of network activation in the prefrontal and parietal areas in both tasks.It implies a top-down predictive mechanism along with the bottom-up progression.This bidirectional interaction could be potentially implemented through frequency-specific synchronization and desynchronization between functional-specific cortical regions,laying the foundation of the speech chain system with common neural substrates. 展开更多
关键词 speech perception and production electroencephalography(EEG)techniques brain network dynamics predictive coding frequency multiplexing
暂未订购 下载PDF
已选0条
引用分析
参考文献
引证文献
 统计分析
检索结果
已选文献
上一页 1 2 … 250 下一页 到第 页
在线咨询 使用帮助 返回顶部 意见反馈