检索结果-内蒙古大学图书馆

Integrating Time-Frequency Domain Shallow and Deep Features for Speech-EEG Match-Mismatch of Auditory Attention Decoding

引用

Journal of Shanghai Jiaotong University (Science) 2025年 1-7页

作者： Zhang, Yubang Zhu, Qiushi Xu, Qingtian Zhang, Jie National Engineering Research Center of Speech and Language Information Processing University of Science and Technology of China Hefei230026 China

Electroencephalogram (EEG) signals provide an important pathway to reflect brain activations, from which auditory attention clues of the listener can be decoded, termed as auditory attention decoding (AAD). However, existing AAD methods primarily rely on temporal or frequency features of audio and shallow features of EEG. In this work, we propose a new model fusion based AAD method with residual dilated convolution blocks, which considers both shallow and deep attention mechanisms as well as time-frequency domain features. Besides, EEG data from different utterances are mixed with the selected EEG segment for augmentation to increase the sample diversity. The effectiveness of our approach is verified by the match-mismatch task of ICASSP2024 Auditory EEG Challenge, which is a typical example of AAD. It performs much better than the baseline and state-of-the-art methods. © Shanghai Jiao Tong University 2025.

关键词： Electroencephalography

来源：评论

学校读者我要写书评

暂无评论

Benchmarking the Performance of Pre-trained LLMs across Urdu NLP Tasks 1

Benchmarking the Performance of Pre-trained LLMs across Urdu...

引用

1st Workshop on Challenges in Processing South Asian languages, CHiPSAL 2025, co-located with the 31st International Conference on Computational Linguistics, COLING 2025

作者： Tahir, Munief Hassan Shams, Sana Fiaz, Layba Adeeba, Farah Hussain, Sarmad Center for Language Engineering Al-Khawarizmi Institute of Computer Science University of Engineering and Technology Lahore Pakistan

ISBN: (纸本)9798891762015

Large language Models (LLMs) pre-trained on multilingual data have revolutionized natural language processing research, by transitioning from languages and task specific model pipelines to a single model adapted on a variety of tasks. However majority of existing multilingual NLP benchmarks for LLMs provide evaluation data in only few languages with little linguistic diversity. In addition these benchmarks lack quality assessment against the respective state-of the art models. This study presents an in-depth examination of 7 prominent LLMs: GPT-3.5-turbo, Llama 2-7B-Chat, Llama 3.1-8B, Bloomz 3B, Bloomz 7B1, Ministral-8B and Whisper (Large, medium and small variant) across 17 tasks using 22 datasets, 13.8 hours of speech, in a zero-shot setting, and their performance against state-of-the-art (SOTA) models, has been compared and analyzed. Our experiments show that SOTA models currently outperform encoder-decoder models in majority of Urdu NLP tasks under zero-shot settings. However, comparing Llama 3.1-8B over prior version Llama 2-7B-Chat, we can deduce that with improved language coverage, LLMs can surpass these SOTA models. Our results emphasize that models with fewer parameters but richer language-specific data, like Llama 3.1-8B, often outperform larger models with lower language diversity, such as GPT-3.5, in several tasks. © 2025 International Committee on Computational Linguistics.

关键词： Computational linguistics

来源：评论

学校读者我要写书评

暂无评论

Benchmarking Whisper for Low-Resource Speech Recognition: An N-Shot Evaluation on Pashto, Punjabi, and Urdu 1

Benchmarking Whisper for Low-Resource Speech Recognition: An...

引用

1st Workshop on Challenges in Processing South Asian languages, CHiPSAL 2025, co-located with the 31st International Conference on Computational Linguistics, COLING 2025

作者： Sehar, Najm Ul Khalid, Ayesha Adeeba, Farah Hussain, Sarmad Center for Language Engineering KICS UET Lahore Pakistan Department of Computer Science UET Lahore Pakistan

ISBN: (纸本)9798891762015

Whisper, a large-scale multilingual model, has demonstrated strong performance in speech recognition benchmarks, but its effectiveness on low-resource languages remains under-explored. This paper evaluates Whisper’s performance on Pashto, Punjabi, and Urdu, three underrepresented languages. While Automatic Speech Recognition (ASR) has advanced for widely spoken languages, low-resource languages still face challenges due to limited data. Whisper’s zero-shot performance was benchmarked and then its small variant was fine-tuned to improve transcription accuracy. Significant reductions in Word Error Rate (WER) were achieved through few-shot fine-tuning, which helped the model better handle challenges such as complex phonetic structures, compared to zero-shot performance. This study contributes to improving multilingual ASR for low-resource languages and highlights Whisper’s adaptability and potential for further enhancement. © 2025 International Committee on Computational Linguistics.

关键词： Computational linguistics

来源：评论

学校读者我要写书评

暂无评论

Bridging the Bandwidth Gap: A Mixed Band Telephonic Urdu ASR Approach with Domain Adaptation for Banking Applications 1

Bridging the Bandwidth Gap: A Mixed Band Telephonic Urdu ASR...

引用

1st Workshop on Challenges in Processing South Asian languages, CHiPSAL 2025, co-located with the 31st International Conference on Computational Linguistics, COLING 2025

作者： Khalid, Ayesha Adeeba, Farah Sehar, Najm Ul Hussain, Sarmad Center for Language Engineering Al-Khawarizmi Institute of Computer Science University of Engineering and Technology Lahore Pakistan Department of Computer Science University of Engineering and Technology New Campus Pakistan

ISBN: (纸本)9798891762015

The accuracy of Automatic Speech Recognition (ASR) systems is influenced by the quality and context of speech signals, particularly in telephonic environments prone to errors like channel drops and noise, leading to higher Word Error Rates (WER). This paper presents the development of a large vocabulary Urdu ASR system for telephonic speech, based on a corpus of 445 speakers from diverse domains. The corpus, annotated at the sentence level, is used to train and evaluate GMM-HMM and chain Time-Delay Neural Network (TDNN) models on a 10-hour test set. Results show that the TDNN model outperforms GMM-HMM. Mixing narrowband and wideband speech further reduces WER. The test sets are also evaluated for the pre-trained model Whisper for performance comparison. Additionally, system adaptation for the banking domain with a specialized lexicon and language model demonstrates the system’s potential for domain-specific applications. © 2025 International Committee on Computational Linguistics.

关键词： Computational linguistics

来源：评论

学校读者我要写书评

暂无评论

Prototype based Masked Audio Model for Self-Supervised Learning of Sound Event Detection

Prototype based Masked Audio Model for Self-Supervised Learn...

引用

2025 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2025

作者： Cai, Pengfei Song, Yan Jiang, Nan Gu, Qing McLoughlin, Ian National Engineering Research Center of Speech and Language Information Processing University of Science and Technology of China China ICT Cluster Singapore Institute of Technology Singapore

ISBN: (纸本)9798350368741

A significant challenge in sound event detection (SED) is the effective utilization of unlabeled data, given the limited availability of labeled data due to high annotation costs. Semi-supervised algorithms rely on labeled data to learn from unlabeled data, and the performance is constrained by the quality and size of the former. In this paper, we introduce the Prototype based Masked Audio Model (PMAM) algorithm for self-supervised representation learning in SED, to better exploit unlabeled data. Specifically, semantically rich frame-level pseudo labels are constructed from a Gaussian mixture model (GMM) based prototypical distribution modeling. These pseudo labels supervise the leaning of a Transformer-based masked audio model, in which binary cross-entropy loss is employed instead of the widely used InfoNCE loss, to provide independent loss contributions from different prototypes, which is important in real scenarios in which multiple labels may apply to unsupervised data frames. A final stage of fine-tuning with just a small amount of labeled data yields a very high performing SED model. On like-for-like tests using the DESED task, our method achieves a PSDS1 score of 62.5%, surpassing current state-of-the-art models and demonstrating the superiority of the proposed technique. © 2025 IEEE.

关键词： masked audio model prototype self-supervised learning sound event detection

来源：评论

学校读者我要写书评

暂无评论

The USTC System for EEG-Music Emotion Recognition Challenge

The USTC System for EEG-Music Emotion Recognition Challenge

引用

2025 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2025

作者： Chen, Jiaxin Wang, Yiming Liu, Yin-Long Feng, Rui Yuan, Jiahong Ling, Zhen-Hua National Engineering Research Center of Speech and Language Information Processing University of Science and Technology of China Hefei China Interdisciplinary Research Center for Linguistic Sciences University of Science and Technology of China Hefei China

ISBN: (纸本)9798350368741

This paper presents the Neural Harmony team's submission to Task 1 (Person Identification) of the ICASSP 2025 EEG-Music Emotion Recognition Challenge, which aims to identify the subject from a given EEG segment. To enhance performance, we propose a novel architecture incorporating the Multiscale ConvBlock and integrating attention mechanisms with convolutional networks. We also reprocessed the data and trained multiple models with different train-validation splits, which were ensembled during testing to further improve robustness. Our final results on the test data exceed the challenge baseline, achieving 100% accuracy in Person Identification. Additionally, unseen subjects were introduced to evaluate the model's generalization ability, and the results confirm the model's strong adaptability to new subjects. © 2025 IEEE.

关键词： attention convolutional network EEG Person Identification

来源：评论

学校读者我要写书评

暂无评论

A survey on textual emotion cause extraction in social networks

引用

Digital Communications and Networks 2025年第2期11卷 524-536页

作者： Sancheng Peng Lihong Cao Guojun Wang Zhouhao Ouyang Yongmei Zhou Shui Yu Laboratory of Language Engineering and Computing Center for Linguistics and Applied Linguistics Guangdong University of Foreign Studies School of English Education Guangdong University of Foreign Studies School of Computer Science and Cyber Engineering Guangzhou University South China University of Technology School of Information Science and Technology Guangdong University of Foreign Studies School of Computer Science University of Technology Sydney

With the rapid development of web technology, Social Networks(SNs) have become one of the most popular platforms for users to exchange views and to express their emotions. More and more people are used to commenting on a certain hot spot in SNs, resulting in a large amount of texts containing emotions. Textual Emotion Cause Extraction(TECE) aims to automatically extract causes for a certain emotion in texts, which is an important research issue in natural language processing. It is different from the previous tasks of emotion recognition and emotion classification. In addition, it is not limited to the shallow-level emotion classification of text, but to trace the emotion source. In this paper, we provide a survey for TECE. First, we introduce the development process and classification of TECE. Then, we discuss the existing methods and key factors for TECE. Finally, we enumerate the challenges and developing trend for TECE.

关键词： Text Emotion Emotion cause Machine learning Deep learning

来源：评论

学校读者我要写书评

暂无评论

UrduLLaMA 1.0: Dataset Curation, Preprocessing, and Evaluation in Low-Resource Settings

arXiv

引用

arXiv 2025年

作者： Fiaz, Layba Tahir, Munief Hassan Shams, Sana Hussain, Sarmad Center for Language Engineering Al-Khawarizmi Institute of Computer Science University of Engineering and Technology Lahore Pakistan

Multilingual Large language Models (LLMs) often provide suboptimal performance on low-resource languages like Urdu. This paper introduces UrduLLaMA 1.0, a model derived from the open-source Llama-3.1-8B-Instruct architecture and continually pre-trained on 128 million Urdu tokens, capturing the rich diversity of the language. To enhance instruction-following and translation capabilities, we leverage Low-Rank Adaptation (LoRA) to fine tune the model on 41,000 Urdu instructions and approximately 50,000 English-Urdu translation pairs. Evaluation across three machine translation datasets demonstrates significant performance improvements compared to state-of-the-art (SOTA) models, establishing a new benchmark for Urdu LLMs. These findings underscore the potential of targeted adaptation strategies with limited data and computational resources to address the unique challenges of low-resource languages. © 2025, CC BY-NC-SA.

关键词： Open systems

来源：评论

学校读者我要写书评

暂无评论

An Investigation on the Impact of Corrosion on the Bond Mechanism of Strand Assemblies 9th

An Investigation on the Impact of Corrosion on the Bond Mech...

引用

9th International Conference on Civil engineering and Materials Science, ICCEMS 2024

作者： Hsu, Shih-Tsung Li, Chung-Pin Hu, Wen-Chi Zhang, Haojie Liao, Jr-Hao Tsai, Yueh-Lun Department of Civil and Construction Engineering Chaoyang University of Technology Taichung City Taiwan Department of Construction Engineering Chaoyang University of Technology Taichung City Taiwan Language Center Chaoyang University of Technology Taichung City Taiwan Department of Construction Engineering Taiwan University of Science and Technology Taipei Taiwan

ISBN: (纸本)9789819615735

Anchors are a building material. Material corrosion is one of the main reasons for anchor failure or function degradation. This paper discusses the impact of corrosion on the bond mechanism of the strand-cement system using the pull-out test of corroded steel strands. According to the test results, the curve of the bond stress-slip relationship of the corroded strand assembly is drawn. The bond mechanisms of corroded and non-corroded steel strands are compared and analyzed. In the experimental group, sodium chloride of 5% solution is added to simulate the actual effect of the corrosion of ground anchors. According to the test results, chloride ions accelerate the development of corrosion as time passes. The bond stress-slip curve can be divided into four stages: elastic stage, slip stage, lower stage, and residual stage. When the strand is slightly rusted, the ultimate bond strength of the corroded strand is higher than that of the non-corroded strand. With the deepening of corrosion on strands, the ultimate bond strength decreases gradually. Two prediction equations are established for the bond strength of strand assemblies under different conditions, and predicted values agree with those tested results. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Steel corrosion

来源：评论

学校读者我要写书评

暂无评论

Voice Attribute Editing With Text Prompt

IEEE Transactions on Audio, Speech and Language Processing

引用

IEEE Transactions on Audio, Speech and language Processing 2025年 33卷 1641-1652页

作者： Zheng-Yan Sheng Li-Juan Liu Yang Ai Jia Pan Zhen-Hua Ling National Engineering Research Center of Speech and Language Information Processing University of Science and Technology of China Hefei China iFLYTEK Research Hefei China

Despite recent advancements in speech generation with text prompt providing control over speech style, voice attributes in synthesized speech remain elusive and challenging to control. This paper introduces a novel task: voice attribute editing with text prompt, with the goal of making relative modifications to voice attributes according to the actions described in the text prompt. To solve this task, VoxEditor, an end-to-end generative model, is proposed. In VoxEditor, addressing the insufficiency of text prompt, a residual memory (ResMem) module is designed, that efficiently maps voice attributes and these descriptors into the shared feature space. Additionally, the ResMem module is enhanced with a voice attribute degree prediction (VADP) module to align voice attributes with corresponding descriptors, addressing the imprecision of text prompt caused by non-quantitative descriptions of voice attributes. We also establish the open-source VCTK-RVA dataset, which leads the way in manual annotations detailing voice characteristic differences among different speakers. Extensive experiments demonstrate the effectiveness and generalizability of our proposed method in terms of both objective and subjective metrics.

关键词： Annotations Training Speech Magnetic domains Manuals Magnetic resonance imaging Linguistics Generative adversarial networks Face recognition Electronic mail

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：