Zobrazeno 1 - 10
of 10 146
pro vyhledávání: '"Wei, Tao"'
Autor:
Zhuang, Ziyang, Miao, Chenfeng, Zou, Kun, Gong, Shuai, Fang, Ming, Wei, Tao, Li, Zijian, Hu, Wei, Wang, Shaojun, Xiao, Jing
Non-autoregressive (NAR) automatic speech recognition (ASR) models predict tokens independently and simultaneously, bringing high inference speed. However, there is still a gap in the accuracy of the NAR models compared to the autoregressive (AR) mod
Externí odkaz:
http://arxiv.org/abs/2406.08835
Autor:
Wang, Xiaowang, Gadre, Chaitanya A., Yang, Runqing, Zou, Wanjuan, Bin, Xing, Addiego, Christopher, Aoki, Toshihiro, Quan, Yujie, Peng, Wei-Tao, Huang, Yifeng, Du, Chaojie, Xu, Mingjie, Yan, Xingxu, Wu, Ruqian, Ong, Shyue Ping, Liao, Bolin, Cao, Penghui, Pan, Xiaoqing
Manipulating thermal properties in materials has been of fundamental importance for advancing innovative technologies. Heat carriers such as phonons are impeded by breaking crystal symmetry or periodicity. Notable methods of impeding the phonon propa
Externí odkaz:
http://arxiv.org/abs/2405.07464
Autor:
Du, Long-Kun, Hu, Chenyu, Liu, Shuang, Deng, Chenjin, Wang, Chaoran, Bo, Zunwang, Chen, Mingliang, Liu, Wei-Tao, Han, Shensheng
Computational imaging~(CI) has been attracting a lot of interest in recent years for its superiority over traditional imaging in various applications. In CI systems, information is generally acquired in an encoded form and subsequently decoded via pr
Externí odkaz:
http://arxiv.org/abs/2401.00032
Currently, sample-specific backdoor attacks (SSBAs) are the most advanced and malicious methods since they can easily circumvent most of the current backdoor defenses. In this paper, we reveal that SSBAs are not sufficiently stealthy due to their poi
Externí odkaz:
http://arxiv.org/abs/2312.04584
Despite being one of the oldest and most widely-used turbulence models in engineering CFD, the k-{\omega} model has not been fully understood theoretically because of its high non-linearity and complex model parameter setting. Here, a multi-layer ana
Externí odkaz:
http://arxiv.org/abs/2311.05454
The pressure gradient has a significant influence on the mean-flow properties of turbulent boundary layers. Conventional analytical studies on PG turbulent boundary layers are conducted separately for the viscous sub-layer and overlap layer only, wit
Externí odkaz:
http://arxiv.org/abs/2311.01903
Data augmentation is vital to the generalization ability and robustness of deep neural networks (DNNs) models. Existing augmentation methods for speaker verification manipulate the raw signal, which are time-consuming and the augmented samples lack d
Externí odkaz:
http://arxiv.org/abs/2310.12111
Autor:
Wu, Shuang, Fei, Zaiyao, Sun, Zeyuan, Yi, Yangfan, Xia, Wei, Yan, Dayu, Guo, Yanfeng, Shi, Youguo, Yan, Jiaqiang, Cobden, David H., Liu, Wei-Tao, Xu, Xiaodong, Wu, Shiwei
Topological properties in quantum materials are often governed by symmetry and tuned by crystal structure and external fields, and hence symmetry-sensitive nonlinear optical measurements in a magnetic field are a valuable probe. Here we report nonlin
Externí odkaz:
http://arxiv.org/abs/2309.09512
Autor:
Dong, Ye, Lu, Wen-jie, Zheng, Yancheng, Wu, Haoqi, Zhao, Derun, Tan, Jin, Huang, Zhicong, Hong, Cheng, Wei, Tao, Chen, Wenguang
With ChatGPT as a representative, tons of companies have began to provide services based on large Transformers models. However, using such a service inevitably leak users' prompts to the model provider. Previous studies have studied secure inference
Externí odkaz:
http://arxiv.org/abs/2307.12533
Autor:
Peng, Longkang, Wei, Tao, Chen, Xuehong, Chen, Xiaobei, Sun, Rui, Wan, Luoma, Chen, Jin, Zhu, Xiaolin
Convolutional neural networks (ConvNets) have been successfully applied to satellite image scene classification. Human-labeled training datasets are essential for ConvNets to perform accurate classification. Errors in human-annotated training dataset
Externí odkaz:
http://arxiv.org/abs/2305.12106