Accurate lung nodule segmentation from Computed Tomog-raphy (CT) images is crucial to the analysis and diagnosis of lung diseases such as COVID-19 and lung cancer. How-ever, due to the variety of lung nodules and the lack of high-quality labeling, accurate lung nodule segmentation is still a challenging problem. In this paper, we propose a novel paradigm including an automatic accurate annotation pipeline and a segmentation network for this task. First, we introduce a new segmentation mask representation named Soft Mask which has richer and more accurate edge details description and better visualization, and we design a universal automatic Soft Mask annotation pipeline to deal with different datasets. Besides, we provide a new challenging lung nodules segmen-tation dataset with traditional binarized masks and our soft masks for further studies. Second, we propose an effective network called SoftGAN that includes an improved back-bone and an adversarial training framework with Soft Mask, in order to improve the performance of accurate lung nodules segmentation. Extensive experiments validate that our Soft-GAN outperforms the state-of-the-art methods for accurate lung nodule segmentation. [Datasetrelease]

Cloud computing is a new software paradigm for resource integration and sharing in the open, dynamic and autonomous network environment. For large-scale complex software system on the dynamic evolution of the software architecture, taking service as the basic unit, the user can select service components and configuration online, can release of application form, and the ability to modify the published application. Current service dependencies including user application software architecture of logic dependency and dependency of evolution in the history of evolution, through the analysis and optimization of clustering in the service dependencies and reconstruction scheme of the user application software architecture is given.

Quality of Service (QoS) has been widely used for personalized Web service recommendation. Since QoS information usually cannot be predetermined, how to make personalized QoS prediction precisely becomes a challenge of Web service recommendation. Time series forecasting and collaborative filtering are two mainstream technologies for QoS prediction. However, on one hand, existing time series forecasting approaches based on Auto Regressive Integrated Moving Average (ARIMA) models do not take the latest observation as a feedback to revise forecasts. Moreover, they only focus on predicting future QoS values for each individual Web service. Service users' personalized factors are not taken into account. On the other hand, collaborative filtering facilitates user-side personalized QoS evaluation, but rarely precisely models the temporal dynamics of QoS values. To address the limitations of existing QoS prediction methods, this paper proposes a novel personalized QoS prediction approach considering both the temporal dynamics of QoS attributes and the personalized factors of service users. Our approach seamlessly combines collaborative filtering with improved time series forecasting which uses Kalman filtering to compensate for shortcomings of ARIMA models. Finally, the experimental results show that the proposed approach can improve the accuracy of personalized QoS prediction significantly.

With the incessant growth of web services on the Internet, how to design effective web service recommendation technologies based on Quality of Service (QoS) is becoming more and more important. Web service recommendation can relieve users from tough work on service selection and improve the efficiency of developing service-oriented applications. Neighborhood-based collaborative filtering has been widely used for web service recommendation, in which similarity measurement and QoS prediction are two key issues. However, traditional similarity models and QoS prediction methods rarely consider the influence of time information, which is an important factor affecting the QoS performance of web services. Furthermore, it is difficult for the existing similarity models to capture the actual relationships between users or services due to data sparsity. The two shortcomings seriously devalue the performance of neighborhood-based collaborative filtering. In this paper, the authors propose an improved time-aware collaborative filtering approach for high-quality web service recommendation. Our approach integrates time information into both similarity measurement and QoS prediction. Additionally, in order to alleviate the data sparsity problem, a hybrid personalized random walk algorithm is designed to infer indirect user similarities and service similarities. Finally, a series of experiments are provided to validate the effectiveness of our approach.

In response to the challenges for application development, deployment, running and maintenance in an open, dynamic Internet platform, the concept of Internetware came into being. It is a new software architecture, which is flexible to evolution, continuously reactive and adaptive to multiple targets. Internetware is able to perceive dynamic changes from internal or external environment, and then do some static adjustment and dynamic evolution according to functional and nonfunctional index to keep a high customer satisfaction. Concept of SOA (Service-Oriented Architecture) is a business-IT-aligned approach in which applications rely on available services to facilitate business processes. Service requestor can flexibility invoke reusable services provided by service provider. However, if your business change, your application consisting of any services need to add, replace, delete any services. Therefore, you have to modify the configuration file or code. Why we don't apply the mature evolution and self-adaption in Internetware to SOA. Then, changes perceived can drive the changes of application based on SOA and our work is to guide these changes in high-level or business-level rather than work in coding-level. Therefore, this paper presents a CCO model binded the SCA(Service Component Architecture) to provide interfaces, which can be invoked by the evolution system or human being's guidance. Then, the application based on SOA can self-adapt the changes. In order to achieve our ideas, we develop an modeling tool based on C/S, named CCOtool, to present the mechanism.

With the incessant growth of Web services on the Internet, designing effective Web service recommendation technologies based on Quality of Service (QoS) is becoming more and more important. Neighborhood-based Collaborative Filtering has been widely used for Web service recommendation, in which similarity measurement and QoS prediction are two key steps. However, traditional similarity models and QoS prediction methods rarely consider the influence of time information, which is an important factor affecting the QoS of Web services. Furthermore, traditional similarity models fail to capture the actual relationships between users or services due to data sparsity. These shortcomings seriously devalue the performance of neighborhood-based Collaborative Filtering. In order to make high-quality Web service recommendation, we propose a novel time-aware approach, which integrates time information into both the similarity measurement and the final QoS prediction. Additionally, in order to alleviate the data sparsity problem, a hybrid personalized random walk algorithm is employed to infer more indirect user similarities and service similarities. Finally, we conduct series of experiments to validate the effectiveness of our approaches.

Due to the increase in web services, non-functional or QoS parameters have been widely addressed. QoS-aware service selection is commonly formulated as an optimization problem aimed at achieving the execution plan that optimize the overall QoS parameter of the composed application while satisfying all the constraints on individual QoS parameters and considering all of the possible execution scenarios. In existing work, execution plans commonly assign one service for each task. For composed applications containing exclusive branches, some tasks may appear in multiple execution scenarios. Unfortunately, since one-fit-all services for these tasks considering every execution scenario are not always available, execution plans have to make compromise. As a result, when some users have strict QoS constraints, feasible execution plans may not exist for the composed application but exist for every execution scenario. This problem becomes more important and challenging for self-adaptive systems, due to the overhead of unwanted runtime adaptations. In this paper, we breaking such one-one mapping between services and tasks. We provide methods to divide all execution scenarios a task involved into exclusive groups (i.e. execution path set partitions) and show that a distinct service can be bound to this task for each group without causing any confusion. In order to find the optimal execution plan, we modify existing MILP based approaches to cooperate with our method. We also evaluate our approach experimentally using real dataset.The results show that it is more flexible, robust to QoS constraints, and can possibly improve the overall QoS associated with the optimal execution plan.

This paper proposes a fast morphological algorithm for color image multi-scale segmentation based on the connected sets using the vertex-collapse algorithm. Color morphological operators are defined based on a special full order relation of color vectors in HSV model. Morphological opening-closing operators are implemented using vertex-collapse algorithm to form a nonlinear scale space. Both area and color are considered to calculate the similarity between adjacent regions, and entropy criterion is adopted to control the content contained in the finial scale tree. The experimental result shows that the image is segmented efficiently with perfect shape preserving.