The change rules associated with hot deformation of FGH96 alloy were investigated by isothermal two-pass hot deformation tests in the temperature range 1050–1125°C and at strain rates ranging from 0.001 to 0.1 s...The change rules associated with hot deformation of FGH96 alloy were investigated by isothermal two-pass hot deformation tests in the temperature range 1050–1125°C and at strain rates ranging from 0.001 to 0.1 s^(-1) on a Gleeble 3500 thermo-simulation machine. The results showed that the softening degree of the alloy between passes decreases with increasing temperature and decreasing strain rates. The critical strain of the first-pass is greater than that of the second-pass. The true stress–true strain curves showed that single-peak dynamic recrystallization, multi-peak dynamic recrystallization, and dynamic response occur when the strain rate is 0.1, 0.01, and 0.001 s^(-1), respectively. The alloy contains three different grain structures after hot deformation: partially recrystallized tissue, completely fine recrystallized tissue, coarse-grained grains. The small-angle grain boundaries increase with increasing temperature. Increasing strain rates cause the small-angle grain boundaries to first increase and then decrease.展开更多
Acoustic scene classification(ASC)is a method of recognizing and classifying environments that employ acoustic signals.Various ASC approaches based on deep learning have been developed,with convolutional neural networ...Acoustic scene classification(ASC)is a method of recognizing and classifying environments that employ acoustic signals.Various ASC approaches based on deep learning have been developed,with convolutional neural networks(CNNs)proving to be the most reliable and commonly utilized in ASC systems due to their suitability for constructing lightweight models.When using ASC systems in the real world,model complexity and device robustness are essential considerations.In this paper,we propose a two-pass mobile network for low-complexity classification of the acoustic scene,named TP-MobNet.With inverse residuals and linear bottlenecks,TPMobNet is based on MobileNetV2,and following mobile blocks,coordinate attention and two-pass fusion approaches are utilized.The log-range dependencies and precise position information in feature maps can be trained via coordinate attention.By capturing more diverse feature resolutions at the network’s end sides,two-pass fusions can also train generalization.Also,the model size is reduced by applying weight quantization to the trained model.By adding weight quantization to the trained model,the model size is also lowered.The TAU Urban Acoustic Scenes 2020 Mobile development set was used for all of the experiments.It has been confirmed that the proposed model,with a model size of 219.6 kB,achieves an accuracy of 73.94%.展开更多
针对隐私敏感物联网(IoT)场景中联邦学习面临巨大通信开销和梯度反演隐私泄露风险问题,提出一种名为QPR(Quantization and Pull Reduction)的两阶段联邦学习通信压缩框架。首先,将训练节点通过梯度量化技术压缩本地梯度后上传至服务器,...针对隐私敏感物联网(IoT)场景中联邦学习面临巨大通信开销和梯度反演隐私泄露风险问题,提出一种名为QPR(Quantization and Pull Reduction)的两阶段联邦学习通信压缩框架。首先,将训练节点通过梯度量化技术压缩本地梯度后上传至服务器,以降低梯度传输的开销;其次,引入基于概率阈值的延迟模型下载机制(lazy pulling)以降低模型的同步频率,训练节点以预设概率同步全局模型,而其余迭代中复用本地历史模型;最后,通过严格的理论分析确保QPR在收敛速度上达到与标准无通信压缩联邦学习——联邦平均(FedAvg)算法相同的渐进阶次,且具备随训练节点数增多而线性加速的特性,从而保证系统的可延展性。实验结果表明,QPR在多个基准数据集和机器学习模型上均能显著提升通信效率。以ResNet18模型在CIFAR-10数据集上的训练任务为例,QPR在不损失模型精度的前提下,与无压缩的FedAvg算法相比,最高实现了8.27的通信加速比。展开更多
基金Financial support from the National Natural Science Foundation of China (No. 51471023)the Ministry of Science and Technology of the People’s Republic of China (National 973 Program, No. 2014GB120000)
文摘The change rules associated with hot deformation of FGH96 alloy were investigated by isothermal two-pass hot deformation tests in the temperature range 1050–1125°C and at strain rates ranging from 0.001 to 0.1 s^(-1) on a Gleeble 3500 thermo-simulation machine. The results showed that the softening degree of the alloy between passes decreases with increasing temperature and decreasing strain rates. The critical strain of the first-pass is greater than that of the second-pass. The true stress–true strain curves showed that single-peak dynamic recrystallization, multi-peak dynamic recrystallization, and dynamic response occur when the strain rate is 0.1, 0.01, and 0.001 s^(-1), respectively. The alloy contains three different grain structures after hot deformation: partially recrystallized tissue, completely fine recrystallized tissue, coarse-grained grains. The small-angle grain boundaries increase with increasing temperature. Increasing strain rates cause the small-angle grain boundaries to first increase and then decrease.
基金This work was supported by Institute of Information&communications Technology Planning&Evaluation(IITP)grant funded by the Korea government(MSIT)[No.2021-0-0268,Artificial Intelligence Innovation Hub(Artificial Intelligence Institute,Seoul National University)]。
文摘Acoustic scene classification(ASC)is a method of recognizing and classifying environments that employ acoustic signals.Various ASC approaches based on deep learning have been developed,with convolutional neural networks(CNNs)proving to be the most reliable and commonly utilized in ASC systems due to their suitability for constructing lightweight models.When using ASC systems in the real world,model complexity and device robustness are essential considerations.In this paper,we propose a two-pass mobile network for low-complexity classification of the acoustic scene,named TP-MobNet.With inverse residuals and linear bottlenecks,TPMobNet is based on MobileNetV2,and following mobile blocks,coordinate attention and two-pass fusion approaches are utilized.The log-range dependencies and precise position information in feature maps can be trained via coordinate attention.By capturing more diverse feature resolutions at the network’s end sides,two-pass fusions can also train generalization.Also,the model size is reduced by applying weight quantization to the trained model.By adding weight quantization to the trained model,the model size is also lowered.The TAU Urban Acoustic Scenes 2020 Mobile development set was used for all of the experiments.It has been confirmed that the proposed model,with a model size of 219.6 kB,achieves an accuracy of 73.94%.
文摘针对隐私敏感物联网(IoT)场景中联邦学习面临巨大通信开销和梯度反演隐私泄露风险问题,提出一种名为QPR(Quantization and Pull Reduction)的两阶段联邦学习通信压缩框架。首先,将训练节点通过梯度量化技术压缩本地梯度后上传至服务器,以降低梯度传输的开销;其次,引入基于概率阈值的延迟模型下载机制(lazy pulling)以降低模型的同步频率,训练节点以预设概率同步全局模型,而其余迭代中复用本地历史模型;最后,通过严格的理论分析确保QPR在收敛速度上达到与标准无通信压缩联邦学习——联邦平均(FedAvg)算法相同的渐进阶次,且具备随训练节点数增多而线性加速的特性,从而保证系统的可延展性。实验结果表明,QPR在多个基准数据集和机器学习模型上均能显著提升通信效率。以ResNet18模型在CIFAR-10数据集上的训练任务为例,QPR在不损失模型精度的前提下,与无压缩的FedAvg算法相比,最高实现了8.27的通信加速比。
