Object-based image labeling through learning by example and multi-level segmentation

We propose a method for automatic extraction and labeling of semantically meaningful image objects using “learning by example” and threshold-free multi-level image segmentation. The proposed method scans through images, each of which is pre-segmented into a hierarchical uniformity tree, to seek and label objects that are similar to an example object presented by the user. By representing images with stacks of multi-level segmentation maps, objects can be extracted in the segmentation map level with adequate detail. Experiments have shown that the proposed multi-level image segmentation results in significant reduction in computation complexity for object extraction and labeling (compared to a single fine-level segmentation) by avoiding unnecessary tests of combinations in finer levels. The multi-level segmentation-based approach also achieves better accuracy in detection and labeling of small objects.     

神经网络在系统建模中的应用研究

提出了一种基于BP算法的余弦基神经网络模型，给出了该神经网络算法收敛性条件，研究了该神经网络算法在系统建模中的应用实例。计算机仿真结果表明该网络模型收敛速度快，在系统建模中不仅是有效的而且是优异的。     

基于示例学习的特征空间变换方法

特征空间是人工智能领域中经常用的基本概念之一，人工智能领域中的许多问题可以可以通过特征空间变换的方法化简和求解。文中提出了一种基于示例学习的特征空间变换方法。     

CLiP限流保护器在中压配电系统中的应用初探

CLiP限流保护器融合了传统限流熔断器的所有优点,是一种一次分断、快速动作的中压故障保护装置。在中压系统中CLiP对有效限制短路电流、减少电能损耗、提高供电质量等,具有良好的作用。文章扼要论述了CLiP限流保护器(简称CLiP)的工作原理和应用实例以及电抗器能耗费用的计算方法。     

有色金属高效反向挤压技术

介绍了高效反向挤压的基本原理及工艺特征与优越性，并在此基础上，分析了高效反向挤压在有色金属中的应用，例举了5MN和25MN高效反向挤压机的结构及生产工艺。     

基于生成对抗网络的小样本数据生成技术研究

基于数据驱动的深度学习技术成为新一代智能电网的应用趋势,该技术对电网中有标注训练数据的量级提出更高的要求。为了获取更多有标注的智能电网样本数据,文章提出了一种基于改进的生成对抗网络（generative adversarial network,GAN）的训练样本生成算法。该方法通过交替训练改进GAN的生成模型与判别模型,无需先验知识的指导,自主学习原始样本的分布规律,生成新的数据样本。然后采用人工神经网络作为基础分类器,计算样本分类的准确率,检验生成样本的有效性。实验表明,改进GAN模型可以有效学习样本的分布规律,提升谐波分类的准确率,该方法同时具有良好的抗噪性和泛化性,对深度学习技术在智能电网中的深入发展具有重要意义。     

ROBUST CONTROL OF GAS TUNGSTEN ARC WELDING SYSTEM

The robust control law for gas tungsten arc welding dynamic process, which is a typical sampled-data system and full of uncertainties, is presented. By using the Lyapunov second method, the robust control and robust optimal control for a class of sampled-data systems whose underlying continuous-time systems are subjected to structured uncertainties are discussed in time-domain. As a result, some sufficient conditions of robust stability and the corresponding robust control laws are derived. All these results are designed by solving a class of linear matrix inequalities (LMIs) and a class of dynamic optimization problem with LMIs constraints respectively. An example adapted under some experimental conditions in the dynamic process of gas tungsten arc welding system in which the controlled variable is the backside width of weld pool and controlling variable pulse duty ratio, is worked out to illustrate the proposed results, it is shown that the sampling period is the crucial design parameter.     

Unpaired image to image transformation via informative coupled generative adversarial networks

We consider image transformation problems, and the objective is to translate images from a source domain to a target one. The problem is challenging since it is difficult to preserve the key properties of the source images, and to make the details of target being as distinguishable as possible. To solve this problem, we propose an informative coupled generative adversarial networks (ICoGAN). For each domain, an adversarial generator-and-discriminator network is constructed. Basically, we make an approximately-shared latent space assumption by a mutual information mechanism, which enables the algorithm to learn representations of both domains in unsupervised setting, and to transform the key properties of images from source to target.Moreover, to further enhance the performance, a weightsharing constraint between two subnetworks, and different level perceptual losses extracted from the intermediate layers of the networks are combined. With quantitative and visual results presented on the tasks of edge to photo transformation, face attribute transfer, and image inpainting, we demonstrate the ICo- GAN’s effectiveness, as compared with other state-of-the-art algorithms.     

融合环境特征与改进YOLOv4的安全帽佩戴检测

目的 在施工现场，安全帽是最为常见和实用的个人防护用具，能够有效防止和减轻意外带来的头部伤害。但在施工现场的安全帽佩戴检测任务中，经常出现难以检测到小目标，或因为复杂多变的环境因素导致检测准确率降低等情况。针对这些问题，提出一种融合环境特征与改进YOLOv4（you only look once version 4）的安全帽佩戴检测方法。方法 为补充卷积池化等过程中丢失的特征，在保证YOLOv4得到的3种不同大小的输出特征图与原图经过特征提取得到的特征图感受野一致的情况下，将两者相加，融合高低层特征，捕捉更多细节信息；对融合后的特征图采用3×3卷积操作，以减小特征图融合后的混叠效应，保证特征稳定性；为适应施工现场的各种环境，利用多种数据增强方式进行环境模拟，并采用对抗训练方法增强模型的泛化能力和鲁棒性。结果 提出的改进YOLOv4方法在开源安全帽佩戴检测数据集（safety helmet wearing dataset，SHWD）上进行测试，平均精度均值（mean average precision，mAP）达到91.55%，较当前流行的几种目标检测算法性能有所提升，其中相比于YOLOv4，mAP提高了5.2%。此外，改进YOLOv4方法在融合环境特征进行数据增强后，mAP提高了4.27%，在各种真实环境条件下进行测试时都有较稳定的表现。结论 提出的融合环境特征与改进YOLOv4的安全帽佩戴检测方法，以改进模型和数据增强的方式提升模型准确率、泛化能力和鲁棒性，为安全帽佩戴检测提供了有效保障。     

Preserving global features of fluid animation from a single image using video examples

We synthesize animations from a single image by transferring fluid motion of a video example globally.Given a target image of a fluid scene,an alpha matte is required to extract the fluid region.Our me...