… on the convolutional neural network (CNN) are AlexNet, … simple backgrounds, when
transferring those methods from … it an ideal model for use in embedded devices. In future work, it is …

… The process of transferring such knowledge requires artificial … neural network and its latest
evolution, deep neural network. … array of input values, and in our case, the pixels of an image. …

… During the movement of the robot,the robot realizes posi⁃ … based on deep learning for
convolutional neural networks. … trained convolutional neural network,uses the convolution feature …

… 对于UC-21 和 AID-30 数据集,非残差网络VGG-16 的分类性能较低, 而… Training deep
convolutional neural networks for land-cover … Learning and transferring convolutional neural …

… of the deep neural network in the past 70 years, the mainstream deep learning model and …
的前提下,深度压缩对于AlexNet 可 以减少35 倍的内存占用,对于VGG-16 的内存占用 可以减少49 …

… 和卷积神经网络 (Convolutional Neural Network,CNN)的相关… ,庄晏榕等[123]使用简化 后的
AlexNet 对母猪双耳竖立行为识别. … behavior recognition based on motion and image analyses[…

… 为MindSpore 深度学习框架训练VGG-16模型时部分算子权重,输出与工作区域的内存占用. 不同
… moDNN: Memory optimal deep neural network training on graphics processing units. IEEE …

针对K-均值算法的差异与缺点, 对初始值敏感, 易于落入局部最优解, 差异进化算法具有强大的
全局收敛能力和鲁棒性, 但其收敛速度较慢. 鉴于上述问题和缺陷, 论文首先详细介绍了进化算法…