Results 31 to 40 of about 815 (137)
Dianci bileiqi, 2022 近年来，基于深度学习的图像识别技术广泛应用于电力巡检图像分析领域。针对电力巡检图像中绝缘子定位算法准确性受到训练图像数量和质量限制的问题，提出一种基于风格迁移算法和图像清洗的电力巡检图像样本扩充方法。首先，提出一种基于风格迁移算法的目标检测数据集扩充方法，通过调整风格迁移算法中风格特征和内容特征占比的系数，生成同时包含绝缘子图像内容特征和风格图像风格特征的风格化图像，并通过复用风格化图像对应的原图像的标注文件，完成风格化图像中绝缘子位置和类别的标注。其次，提出一种基于改进LeNet ...
尤振飞 +6 more
doaj
Bi-directional Pre-trained Network for Single-station Seismic Waveform Analysis [PDF]
The application of machine learning, particularly deep learning methods, is becoming increasingly widespread in seismology, achieving near-human accuracy in tasks such as phase detection and event classification.
Lu LI +4 more
core +1 more source
Wildlife Letters, Volume 3, Issue 3, Page 91-98, September 2025.The edge densities of cropland, developed land, and water bodies (panel a) predict the spatial probability of mammalian predation on wild turkeys (panel b) using Mahalanobis distance factor analysis in Quitman County, Mississippi, United States. The three habitat fragmentation variables are represented by three arrows, respectively.
Guiming Wang +2 more
wiley +1 more source
Fault diagnosis method of rolling bearing of mine main fan based on transfer learning [PDF]
The condition monitoring and fault diagnosis of the rolling bearings of the main fan in the mine are significant to the safety of coal mine production. The existing fault diagnosis methods of rolling bearing have the problems of insufficient training and
Guoying MENG, Wei CUI, Xingwei WAN
core +1 more source
Jianzhu Gangjiegou Jinzhan, 2023 基于支持向量机的结构损伤识别方法建立在训练和测试数据同概率分布的假设上，必须对每个结构分别收集训练数据和标签并训练模型以识别损伤。损伤结构的训练数据和标签难以收集，导致支持向量机方法在结构损伤识别中难以实施。提出一种基于领域自适应支持向量机的跨域损伤识别方法，将在有损伤标签结构（源结构）上训练的支持向量机推广到无损伤标签结构（目标结构）的损伤识别中。该方法首先提取结构的多阶固有频率变化率作为损伤特征构建训练数据；然后，推导最小化边缘分布差异与最小化联合分布差异的迁移学习方法 ...
李佐强 +6 more
doaj
A review of deep learning-based few sample fault diagnosis method for rotating machinery [PDF]
ObjectivesDeep learning has shown great potential in the field of rotating machinery fault diagnosis. Its excellent performance heavily relies on sufficient training samples.
Jun WU +4 more
core +1 more source
Implicit smartphone authentication via multimodal data and transfer autoencoders [PDF]
With the widely used of smartphones, traditional explicit authentication methods (e.g., passwords, fingerprints) are increasingly vulnerable due to their reliance on active user input, making the non-invasive implicit authentication a critical research ...
XU Zi’ang
core +1 more source
智能科学与技术学报, 2021 针对皮肤镜图像数量不充足以及各类疾病之间影像数据不平衡的问题，提出一种融合类加权交叉熵损失函数和分层卷积神经网络的皮肤镜图像分类方法。首先对皮肤镜图像进行色彩恒常化处理，消除环境光源噪声；然后构建基于ResNet 50的分层卷积神经网络，并在迁移学习的基础上分别构建二分类和多分类卷积神经网络模型，根据皮肤镜图像的数量特点设置类加权交叉熵损失函数。实验结果表明，该方法具有较好的分类效果，分类准确率达到了85.94%，与未改进的分类模型ResNet 50相比，测试准确率提高了5.752%。
邵虹, 张鸣坤, 崔文成
doaj
Comparison of Three CNN Models Applied in Bone Age Assessment of Pelvic Radiographs of Adolescents [PDF]
, 2020 Objective To compare the performance of three deep-learning models （VGG19, Inception-V3 and Inception-ResNet-V2） in automatic bone age assessment based on pelvic X-ray radiographs.
PENG Li- qin , WAN Lei , WANG Mao- wen, et al.
core +1 more source
Gaoya dianqi, 2023 绝缘和机械故障是气体绝缘金属封闭开关设备（gas insulated metal-enclosed switchgear,GIS）中占比最大的故障类型，准确的故障诊断和状态评价对保证电力系统安全稳定运行具有重要意义。深度学习方法已成为故障诊断领域的主流，但传统卷积神经网络需要强大的计算资源，在计算能力一般的智能终端设备中难以应用。为此，文中提出了基于轻量级卷积神经网络的GIS绝缘和机械故障诊断方法。首先，采用空间可分离卷积代替传统卷积构造EffNet轻量级卷积神经网络，大幅度降低了模型的计算量；其次 ...
杨为 +4 more
doaj