Results 111 to 120 of about 12,984 (168)
大数据磁盘作为重要的存储介质，一旦出现故障很可能会导致存储数据丢失，给个人及企业带来难以估量的损失。现有磁盘故障预测模型存在不能很好地平衡磁盘数据样本、未充分利用磁盘数据的时序特性等问题。以Backblaze云存储公司公布的真实磁盘数据为研究对象，提出了一种融合注意力机制的卷积神经网络（CNN）和双向门控循环单元（BiGRU）网络的磁盘故障预测模型。在数据预处理方面，采用负采样与焦点损失函数来平衡正负样本，利用CNN进行特征提取，并结合BiGRU网络来有效地处理时序数据。通过融合注意力机制 ...
王艳, 刘亚东, 皮婵娟, 施君豪
doaj
Research on mine underground inspection robot target detection algorithm based on pyramid structure and attention mechanism coupling [PDF]
In recent years, coal mine robots have become a research hotspot in the field of modern coal machine equipment, and the main coal flow transportation system of most coal mines has basically realized continuity, mechanization and automation, which also ...
Jiusheng BAO +5 more
core +1 more source
Operation standardization evaluation method based on improved YOLOv8n for ship equipment disassembly and assembly [PDF]
ObjectivesThe standardization of ship engine room operations is a critical component of ship safety management. Therefore, the practical examination for crew members includes the disassembly and assembly of ship equipment as a key assessment item.
Chao WU +4 more
core +1 more source
Research Review of Deep Learning in Colon Polyp Image Segmentation [PDF]
Colorectal polyp is an abnormal tissue growing in the gastrointestinal tract with the potential to develop into colorectal cancer. Therefore, early detection and removal of colorectal polyps are crucial for preventing colorectal cancer.
LI Guowei, LIU Jing, CAO Hui, JIANG Liang
core +1 more source
AAT model based channel estimation for mmWave massive MIMO systems [PDF]
To solve the problems of temporal correlation and susceptibility to noise in millimeter wave massive MIMO channels, which result in decreased channel estimation accuracy, a novel channel estimation method based on an improved temporal convolutional ...
Liya HUANG +4 more
core +1 more source
Research progress of Ruyi Jinhuang Powder in the treatment of pressure injuries (如意金黄散治疗压力性损伤的研究进展)
中西医结合护理Pressure injury (PI) is a common skin problem that occurs most often in patients who are bedridden or have limited mobility, seriously affecting the quality of life of patients.
ZHU Yuxin (朱雨昕) +2 more
doaj +1 more source
Huli yanjiu, 2005 [目的]探讨施行腰麻-硬膜外联合麻醉（CSEA）无痛分娩产程中,不同时机关闭镇痛泵（PCA）停止注药对第二产程的影响。[方法]选择自愿接受CSEA无痛分娩的孕妇100例,50例在活跃期宫口开大8 cm时关闭镇痛泵停止注药（观察组）,50例于宫口开全后关闭镇痛泵（对照组）,观察两组第二产程时间、自然分娩数、产钳助产和剖宫产数以及第二产程产妇疼痛情况。[结果]观察组第二产程运用腹压有力,第二产程时间短,产钳助产率低,与对照组比较有统计学意义;两组第二产程疼痛情况比较无统计学意义。[结论 ...
卢兰琴, 卢爱金, 唐志红
doaj
Zhejiang Daxue xuebao. Lixue banThe rapid development of the 3D animation, film and television production, and gaming industries has accumulated massive amounts of high-precision 3D human motion data, making efficient management and intelligent retrieval a serious challenge.
YU Congrui(于聪睿) +3 more
doaj +1 more source
Zhendong Ceshi yu Zhenduan针对燃气调压器故障识别中不平衡数据影响模型识别能力的问题，提出一种一维卷积神经网络（one‑dimensional convolutional neural network，简称1D‑CNN）与注意力机制（squeeze‑and‑excitation，简称SE）相结合的改进深度卷积神经网络（SE‑1DCNN）方法。首先，使用一维卷积核提取故障特征；其次，在交替的卷积层后添加SE模块用于通道加权，选择性地保留所需的重要信息特征，并抑制弱相关的特征；最后 ...
doaj +1 more source
基于多头注意力机制的CNN-BiLSTM高海拔多因素输电线路可听噪声预测
Gaoya dianqi为了研究考虑高海拔多环境因素影响下输电线路可听噪声的预测问题，在海拔2 400 m高度点的500 kV同塔双回线路下，搭建了边相外20、30、35 m三处可听噪声观测站，同时利用气象站进行多环境因素指标的数据采集。文中提出了一种基于多头注意力机制（multi-head attention，MHA）的卷积神经网络（convolutional neural network，CNN）—双向长短期记忆网络（bi-directional long short term memory，BiLSTM ...
黄悦华 +4 more
doaj