检索结果-内蒙古大学图书馆

Self-Transriber: Few-Shot Lyrics Transcription With Self-Training

学校读者我要写书评

暂无评论

Self-Transriber: Few-Shot Lyrics Transcription With Self-Tra...

International Conference on Acoustics, Speech, and Signal Processing (ICASSP)

作者： Xiaoxue Gao Xianghu Yue Haizhou Li Department of Electrical and Computer Engineering National University of Singapore Singapore Shenzhen Research Institute of Big Data Shenzhen China School of Data Science The Chinese University of Hong Kong Shenzhen China

The current lyrics transcription approaches heavily rely on supervised learning with labeled data, but such data are scarce and manual labeling of singing is expensive. How to benefit from unlabeled data and alleviate limited data problem have not been explored for lyrics transcription. We propose the first semi-supervised lyrics transcription paradigm, Self-Transcriber, by leveraging on unlabeled data using selftraining with noisy student augmentation. We attempt to demonstrate the possibility of lyrics transcription with a few amount of labeled data. Self-Transcriber generates pseudo labels of the unlabeled singing using teacher model, and augments pseudo-labels to the labeled data for student model update with both self-training and supervised training losses. This work closes the gap between supervised and semi- supervised learning as well as opens doors for few-shot learning of lyrics transcription. Our experiments show that our approach using only 12.7 hours of labeled data achieves competitive performance compared with the supervised approaches trained on 149.1 hours of labeled data for lyrics transcription.

关键词： Training Supervised learning Manuals Signal processing Semisupervised learning data models Acoustics

Algorithmic Collusion in Dynamic Pricing with Deep Reinforcement Learning

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Deng, Shidi Schiffer, Maximilian Bichler, Martin School of Management Technical University of Munich Germany School of Management Munich Data Science Institute Technical University of Munich Germany School of Computation Information and Technology Technical University of Munich Germany

Nowadays, a significant share of the Business-to-Consumer sector is based on online platforms like Amazon and Alibaba and uses Artificial Intelligence for pricing strategies. This has sparked debate on whether pricing algorithms may tacitly collude to set supra-competitive prices without being explicitly designed to do so. Our study addresses these concerns by examining the risk of collusion when Reinforcement Learning algorithms are used to decide on pricing strategies in competitive markets. Prior research in this field focused on Tabular Q-learning (TQL) and led to opposing views on whether learning-based algorithms can lead to supra-competitive prices. Our work contributes to this ongoing discussion by providing a more nuanced numerical study that goes beyond TQL by additionally capturing off- and on-policy Deep Reinforcement Learning (DRL) algorithms. We study multiple Bertrand oligopoly variants and show that algorithmic collusion depends on the algorithm used. In our experiments, TQL exhibits higher collusion and price dispersion phenomena compared to DRL algorithms. We show that the severity of collusion depends not only on the algorithm used but also on the characteristics of the market environment. We further find that Proximal Policy Optimization appears to be less sensitive to collusive outcomes compared to other state-of-the-art DRL algorithms. Copyright © 2024, The Authors. All rights reserved.

关键词： Reinforcement learning

SPIKING-LEAF: A LEARNABLE AUDITORY FRONT-END FOR SPIKING NEURAL NETWORKS

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Song, Zeyang Wu, Jibin Zhang, Malu Shou, Mike Zheng Li, Haizhou National University of Singapore Singapore The Hong Kong Polytechnic University Hong Kong University of Electronic Science and Technology of China China Shenzhen Research Institute of Big Data School of Data Science The Chinese University of Hong Kong Shenzhen China

Brain-inspired spiking neural networks (SNNs) have demonstrated great potential for temporal signal processing. However, their performance in speech processing remains limited due to the lack of an effective auditory front-end. To address this limitation, we introduce Spiking-LEAF, a learnable auditory front-end meticulously designed for SNN-based speech processing. Spiking-LEAF combines a learnable filter bank with a novel two-compartment spiking neuron model called IHC-LIF. The IHC-LIF neurons draw inspiration from the structure of inner hair cells (IHC) and they leverage segregated dendritic and somatic compartments to effectively capture multi-scale temporal dynamics of speech signals. Additionally, the IHC-LIF neurons incorporate the lateral feedback mechanism along with spike regularization loss to enhance spike encoding efficiency. On keyword spotting and speaker identification tasks, the proposed Spiking-LEAF outperforms both SOTA spiking auditory front-ends and conventional real-valued acoustic features in terms of classification accuracy, noise robustness, and encoding efficiency. Copyright © 2023, The Authors. All rights reserved.

关键词： Speech recognition

Summer Atmospheric Water Cycle under the Transition Influence of the Westerly and Summer Monsoon over the Yarlung Zangbo River Basin in the Southern Tibetan Plateau

学校读者我要写书评

暂无评论

Advances in Atmospheric sciences 2024年第5期41卷 830-846页

作者： Qianhui MA Chunyan ZHANG Donghai WANG Zihao PANG School of Atmospheric Sciences Sun Yat-sen UniversityGuangdong Province Key Laboratory for Climate Change and Natural Disaster StudiesKey Laboratory of Tropical Atmosphere-Ocean SystemMinistry of EducationZhuhai 519082China Southern Marine Science and Engineering Guangdong Laboratory(Zhuhai) Zhuhai 519082China Guangdong Meteorological Data Center Guangdong Meteorological ServiceGuangzhou 510080China National Observation and Research Station of Coastal Ecological Environments in Macao Macao Environmental Research InstituteMacao University of Science and TechnologyMacao SAR 999078China National Meteorological Information Center China Meteorological AdministrationBeijing 100081China

This study compares the summer atmospheric water cycle,including moisture sources and consumption,in the upstream,midstream,and downstream regions of the Yarlung Zangbo River Basin in the southern Tibetan *** evolutions of moisture properties under the influence of the westerly and summer southerly monsoon are examined using 5-yr multi-source measurements and ERA5 reanalysis *** that moisture consumption in this study is associated with clouds,precipitation,and diabatic *** to the midstream and downstream regions,the upstream region has less moisture,clouds,and precipitation,where the moisture is brought by the *** early August,the vertical wet advection over this region becomes enhanced and generates more high clouds and *** midstream region has moisture carried by the westerly in June and by the southerly monsoon from July to *** higher vertical wet advection maximum here forms more high clouds,with a precipitation peak in early *** downstream region is mainly affected by the southerly-driven wet *** rich moisture and strong vertical wet advection here produce the most clouds and precipitation among the three regions,with a precipitation peak in late *** height of the maximum moisture condensation is different between the midstream region(325 hPa)and the other two regions(375 hPa),due to the higher upward motion maximum in the midstream *** diabatic heating structures show that stratiform clouds dominate the upstream region,stratiform clouds and deep convection co-exist in the midstream region,and deep convection systems characterize the downstream region.

关键词： Yarlung Zangbo River Basin atmospheric water cycle constrained variational analysis moisture source and consumption

Resampling Factor Estimation via Dual-Stream Convolutional Neural Network

学校读者我要写书评

暂无评论

Computers, Materials & Continua 2021年第1期66卷 647-657页

作者： Shangjun Luo Junwei Luo Wei Lu Yanmei Fang Jinhua Zeng Shaopei Shi Yue Zhang School of Data and Computer Science Guangdong Province Key Laboratory of Information Security TechnologyMinistry of Education Key Laboratory of Machine Intelligence and Advanced ComputingSun Yat-sen UniversityGuangzhou510006China Academy of Forensic Science Shanghai200063China College of Information Science and Technology Jinan UniversityGuangzhou510632China Department of Computer Science University of Massachusetts LowellLowellMA 01854USA

The estimation of image resampling factors is an important problem in image *** all the resampling factor estimation methods,spectrumbased methods are one of the most widely used methods and have attracted a lot of research ***,because of inherent ambiguity,spectrum-based methods fail to discriminate upscale and downscale operations without any prior *** general,the application of resampling leaves detectable traces in both spatial domain and frequency domain of a resampled ***,the resampling process will introduce correlations between neighboring *** this case,a set of periodic pixels that are correlated to their neighbors can be found in a resampled ***,the resampled image has distinct and strong peaks on spectrum while the spectrum of original image has no clear ***,in this paper,we propose a dual-stream convolutional neural network for image resampling factors *** of the two streams is gray stream whose purpose is to extract resampling traces features directly from the rescaled *** other is frequency stream that discovers the differences of spectrum between rescaled and original *** features from two streams are then fused to construct a feature representation including the resampling traces left in spatial and frequency domain,which is later fed into softmax layer for resampling factor *** results show that the proposed method is effective on resampling factor estimation and outperforms some CNN-based methods.

关键词： Image forensics image resampling detection parameter estimation convolutional neural network

A Real-Time Approach to Capture Ambient and Focal Attention in Visual Search 25

学校读者我要写书评

暂无评论

A Real-Time Approach to Capture Ambient and Focal Attention ...

Proceedings of the 2025 Symposium on Eye Tracking Research and Applications

作者： Gavindya Jayawardena Yasith Jayawardana Yasasi Abeysinghe Bhanuka Mahanama Sampath Jayarathna Jacek Gwizdka School of Information The University of Texas at Austin Austin Texas USA Institute of Data Engineering and Science Georgia Institute of Technology Atlanta Georgia USA Department of Computer Science Old Dominion University Norfolk Virginia USA Computer Science Old Dominion University Norfolk Virginia USA

Emotion Rendering for Conversational Speech Synthesis with Heterogeneous Graph-Based Context Modeling

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Liu, Rui Hu, Yifan Ren, Yi Yin, Xiang Li, Haizhou Inner Mongolian University China ByteDance China Shenzhen Research Institute of Big Data School of Data Science The Chinese University of Hong Kong Shenzhen China National University of Singapore Singapore

Conversational Speech Synthesis (CSS) aims to accurately express an utterance with the appropriate prosody and emotional inflection within a conversational setting. While recognising the significance of CSS task, the prior studies have not thoroughly investigated the emotional expressiveness problems due to the scarcity of emotional conversational datasets and the difficulty of stateful emotion modeling. In this paper, we propose a novel emotional CSS model, termed ECSS, that includes two main components: 1) to enhance emotion understanding, we introduce a heterogeneous graph-based emotional context modeling mechanism, which takes the multi-source dialogue history as input to model the dialogue context and learn the emotion cues from the context;2) to achieve emotion rendering, we employ a contrastive learning-based emotion renderer module to infer the accurate emotion style for the target utterance. To address the issue of data scarcity, we meticulously create emotional labels in terms of category and intensity, and annotate additional emotional information on the existing conversational dataset (DailyTalk). Both objective and subjective evaluations suggest that our model outperforms the baseline models in understanding and rendering emotions. These evaluations also underscore the importance of comprehensive emotional annotations. Code and audio samples can be found at: https://***/walkerhyf/ECSS. © 2023, CC BY.

关键词： Speech synthesis

Near-Optimal Resilient Aggregation Rules for Distributed Learning Using 1-Center and 1-Mean Clustering with Outliers

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Yi, Yuhao You, Ronghui Liu, Hong Liu, Changxin Wang, Yuan Lv, Jiancheng College of Computer Science Sichuan University China School of Statistics and Data Science Nankai University China School of Electrical Engineering and Computer Science KTH Royal Institute of Technology Sweden School of Robotics Hunan University China

Byzantine machine learning has garnered considerable attention in light of the unpredictable faults that can occur in large-scale distributed learning systems. The key to secure resilience against Byzantine machines in distributed learning is resilient aggregation mechanisms. Although abundant resilient aggregation rules have been proposed, they are designed in ad-hoc manners, imposing extra barriers on comparing, analyzing, and improving the rules across performance criteria. This paper studies near-optimal aggregation rules using clustering in the presence of outliers. Our outlier-robust clustering approach utilizes geometric properties of the update vectors provided by workers. Our analysis show that constant approximations to the 1-center and 1-mean clustering problems with outliers provide near-optimal resilient aggregators for metric-based criteria, which have been proven to be crucial in the homogeneous and heterogeneous cases respectively. In addition, we discuss two contradicting types of attacks under which no single aggregation rule is guaranteed to improve upon the naive average. Based on the discussion, we propose a two-phase resilient aggregation framework. We run experiments for image classification using a non-convex loss function. The proposed algorithms outperform previously known aggregation rules by a large margin with both homogeneous and heterogeneous data distributions among non-faulty workers. Code and appendix are available at https://***/jerry907/AAAI24-RASHB. Copyright © 2023, The Authors. All rights reserved.

关键词： Machine learning

EEG Sleep Stages Classification using Transformer-based Models: A Review

学校读者我要写书评

暂无评论

EEG Sleep Stages Classification using Transformer-based Mode...

computing Innovation, Intelligence, Technologies and Education (CIITE), International Conference on

作者： Yuin Torng Yew Jia Hong Puah Chow Khuen Chan Norita Mohd Zain Chean Khim Toa Sim Kuan Goh Department of Biomedical Engineering Faculty of Engineering Universiti Malaya School of Energy and Chemical Engineering Xiamen University Malaysia School of Electrical Engineering and Artificial Intelligence Xiamen University Malaysia School of Computing and Data Science Xiamen University Malaysia

ISBN: (数字)9798350377286

ISBN: (纸本)9798350377293

Sleep research is vital for understanding the complexities of human physiology and promoting overall health and well-being. Despite its importance, sleep disorders are widespread and can significantly impact daily functioning and quality of life. A critical aspect of sleep research is the classification of sleep stages, traditionally performed manually but now increasingly automated using computational techniques. However, current classification models face challenges, particularly in accurately distinguishing between sleep stages such as N1, REM, and wakefulness, and in classifying different types of sleep, such as micro sleep and macro sleep. Addressing these research challenges can lead to significant advancements in automated sleep scoring and contribute to the development of clinically deployable sleep diagnostic tools. Recently, there is a growing interest in leveraging advanced deep learning models, especially transformers, for analyzing EEG signals, which are crucial for sleep stage classification due to their temporal dependency. In contrast to convolutional neural networks (CNNs), which have been commonly used for this purpose, transformer architectures offer promise in capturing long-range dependencies in time-series data. Furthermore, the integration of multimodal learning approaches, such as combining EEG signals with other physiological data, holds the potential for improving the accuracy and robustness of sleep stage classification. This review article focuses on transformer-based models for sleep stages classification, and also to clarify the current gaps in sleep classification.

关键词： Deep learning Accuracy Sleep Reviews Computational modeling Computer architecture Transformers Brain modeling Electroencephalography Physiology