Meta-transfer-adjustment learning for few-shot learning

Deep neural network models with strong feature extraction capacity are prone to overfitting and fail to adapt quickly to new tasks with few samples. Gradient-based meta-learning approaches can minimize overfitting and adapt to new tasks fast, but they frequently use shallow neural networks with limited feature extraction capacity. We present a simple and effective approach called Meta-Transfer-Adjustment learning (MTA) in this paper, which enables deep neural networks with powerful feature extraction capabilities to be applied to few-shot scenarios while avoiding overfitting and gaining the capacity for quickly adapting to new tasks via training on numerous tasks. Our presented approach is classified into two major parts, the Feature Adjustment (FA) module, and the Task Adjustment (TA) module. The feature adjustment module (FA) helps the model to make better use of the deep network to improve feature extraction, while the task adjustment module (TA) is utilized for further improve the model’s fast response and generalization capabilities. The proposed model delivers good classification results on the benchmark small sample datasets MiniImageNet and Fewshot-CIFAR100, as proved experimentally.     

Unsupervised learning of multi-task deep variational model

We propose a general deep variational model (reduced version, full version as well as the extension) via a comprehensive fusion approach in this paper. It is able to realize various image tasks in a completely unsupervised way without learning from samples. Technically, it can properly incorporate the CNN based deep image prior (DIP) architecture into the classic variational image processing models. The minimization problem solving strategy is transformed from iteratively minimizing the sub-problem for each variable to automatically minimizing the loss function by learning the generator network parameters. The proposed deep variational (DV) model contributes to the high order image edition and applications such as image restoration, inpainting, decomposition and texture segmentation. Experiments conducted have demonstrated significant advantages of the proposed deep variational model in comparison with several powerful techniques including variational methods and deep learning approaches.     

Mammogram Learning System for Breast Cancer Diagnosis Using Deep Learning SVM

The most common form of cancer for women is breast cancer. Recent advances in medical imaging technologies increase the use of digital mammograms to diagnose breast cancer. Thus, an automated computerized system with high accuracy is needed. In this study, an efficient Deep Learning Architecture (DLA) with a Support Vector Machine (SVM) is designed for breast cancer diagnosis. It combines the ideas from DLA with SVM. The state-of-the-art Visual Geometric Group (VGG) architecture with 16 layers is employed in this study as it uses the small size of 3 × 3 convolution filters that reduces system complexity. The softmax layer in VGG assumes that the training samples belong to exactly only one class, which is not valid in a real situation, such as in medical image diagnosis. To overcome this situation, SVM is employed instead of the softmax layer in VGG. Data augmentation is also employed as DLA usually requires a large number of samples. VGG model with different SVM kernels is built to classify the mammograms. Results show that the VGG-SVM model has good potential for the classification of Mammographic Image Analysis Society (MIAS) database images with an accuracy of 98.67%, sensitivity of 99.32%, and specificity of 98.34%.     

基于对抗深度学习的无人机航拍违建场地识别

对违建场地的检测方法主要是通过人工对无人机航拍视频进行检查,存在检测精度低、识别性能差、工作效率低的问题。提出一种结合空间变换网络与Fast RCNN的生成对抗网络ASTN-Fast RCNN,通过深度学习与无人机航拍视频相结合自动识别检测处在建设初期的违建场地。将空间变换网络作为生成器,生成Fast RCNN目标检测器难以识别的旋转形变样本,并通过目标检测器与生成器的对抗式训练,提高检测器的鲁棒性。实验结果表明,该方法能够有效提高对无人机航拍违建场地的识别性能。     

一种基于YOLOv4-TIA的林业害虫实时检测方法

针对现有基于深度学习的林业昆虫图像检测方法存在检测精度低和检测速度慢的问题,提出一种结合改进PANet结构与三分支注意力机制的目标检测方法YOLOv4-TIA。通过对样本数量较少的昆虫类别进行数据增强,实现样本均衡分布。利用三分支注意力机制改进YOLOv4中的CSPDarkNet53骨干网络,同时通过旋转操作和残差变换建立维度间的依存关系,以提高有效的特征通道权重,在PANet结构上增加将跳跃连接与跨尺度连接相结合的特征融合方式,从而获取更丰富的语义信息和位置信息。在此基础上,采用Focal loss函数优化分类损失,解决正负样本不均衡的问题。实验结果表明,该方法的精确率和召回率分别达到85.9%和91.2%,相比SSD、Faster R-CNN、YOLOv4方法,其在保证检测速度的同时,能够有效提高检测精度,且实现对林业害虫的实时精确监测。     

基于深度学习的监督型典型相关分析

典型相关分析（CCA）是利用综合变量对之间的相关关系反映两组指标之间整体相关性的多元统计方法。传统的CCA方法无法有效利用样本的标签信息,导致准确率降低。将类信息融入到深度学习与CCA相结合的深度典型相关分析中,提出一种监督型降维方法DL-SCCA,用于处理带标签的非线性可分数据。在2个独立的深度神经网络（DNN）结构上,增加1个公共的输出维数与数据集类别数相同的全连接层,并且以softmax函数作为该层的激活函数,输出带有概率意义的编码向量。在此基础上,利用全连接输出与样本标签信息之间的交叉熵对DNN进行训练,获得分类性能较优的低维特征。实验结果表明,该方法采用最近邻分类器和网络本身结构得到的分类准确率分别为98.00%和97.82%,相比CCA、DisCCA、DCCA等方法,能够有效利用样本的标签信息,并且具有较优的分类性能。     

基于深度强化学习的机械臂控制快速训练方法

人工智能在机器人控制中得到广泛应用,机器人控制算法也逐渐从模型驱动转变为数据驱动。深度强化学习算法可在复杂环境中感知并决策,能够解决高维度和连续状态空间下的机械臂控制问题。然而,目前深度强化学习中数据驱动的训练过程非常依赖计算机GPU算力,且训练时间成本较大。提出基于深度强化学习的先简化模型（2D模型）再复杂模型（3D模型）的机械臂控制快速训练方法。采用深度确定性策略梯度算法代替机械臂传统控制算法中的逆运动学解算方法,直接通过数据驱动的训练过程控制机械臂末端到达目标位置,从而减小训练时间成本。同时,对于状态向量和奖励函数形式,使用不同的设置方式。将最终训练得到的算法模型在真实机械臂上进行实现和验证,结果表明,其控制效果达到了分拣物品的应用要求,相比于直接在3D模型中的训练,能够缩短近52%的平均训练时长。     

Optimal Deep Learning Model Enabled Secure UAV Classification forIndustry 4.0

Emerging technologies such as edge computing, Internet of Things (IoT), 5G networks, big data, Artificial Intelligence (AI), and Unmanned Aerial Vehicles (UAVs) empower, Industry 4.0, with a progressive production methodology that shows attention to the interaction between machine and human beings. In the literature, various authors have focused on resolving security problems in UAV communication to provide safety for vital applications. The current research article presents a Circle Search Optimization with Deep Learning Enabled Secure UAV Classification (CSODL-SUAVC) model for Industry 4.0 environment. The suggested CSODL-SUAVC methodology is aimed at accomplishing two core objectives such as secure communication via image steganography and image classification. Primarily, the proposed CSODL-SUAVC method involves the following methods such as Multi-Level Discrete Wavelet Transformation (ML-DWT), CSO-related Optimal Pixel Selection (CSO-OPS), and signcryption-based encryption. The proposed model deploys the CSO-OPS technique to select the optimal pixel points in cover images. The secret images, encrypted by signcryption technique, are embedded into cover images. Besides, the image classification process includes three components namely, Super-Resolution using Convolution Neural Network (SRCNN), Adam optimizer, and softmax classifier. The integration of the CSO-OPS algorithm and Adam optimizer helps in achieving the maximum performance upon UAV communication. The proposed CSODL-SUAVC model was experimentally validated using benchmark datasets and the outcomes were evaluated under distinct aspects. The simulation outcomes established the supreme better performance of the CSODL-SUAVC model over recent approaches.     

应用词法分析方法抽取Web查询接口模式

Deep Web中蕴含着大量高质量的数据,然而只有通过Web查询接口对Web数据库提交查询才能获取这些数据,因此,自动获取Web查询接口模式是实现Web数据库集成的关键.将Web查询接口模式的抽取过程看作一个词法分析的过程,通过构建EGLM-FA(元素分组及标签匹配有限状态自动机)来完成对Web查询接口模式的抽取.首先应用Html呈现引擎将Web查询接口所在页面进行解析,利用查询接口Form中的DOM节点及其坐标信息构建相应的NSS(节点空间结构),之后再将所有的NSS组成NSS列表,将NSS列表作为EGLM-FA的输入,进而抽取出Web查询接口的模式.     

基于世界知识的深网数据源增强分类模型

针对传统词袋方法在深网(Deep Web)数据源分类应用中的局限性,提出一种基于世界知识的Deep Web数据源增强分类模型,通过对外部知识库的主题分析,建立特征映射,构造基于领域概念的辅助分类器,丰富Deep Web查询表单的特征集合。基于Wikipedia百科知识库对真实Web数据进行分类。实验结果证明该模型有效。