Search

Color variations of histological images due to multi-factor hinder the performance of computer-aided diagnosis (CAD) systems. Previous stain normalization methods have achieved excellent results. While in practice, a multidomain stain normalization method is still be needed when more than two color variations exist in dataset. In this paper, we propose a multidomain stain normalization model inspired by MUNIT [1], with the idea of disentangling the representations of content and style. We assume that the latent space of histological images can be decomposed into domain-shared content space and domain-specific style space. The stain normalization aims to transfer the styles cross domains and maintain the contents. In addition, we propose to use the earth mover’s distance(EMD) to evaluate the effectiveness of stain normalization. We evaluate our approach against the state-of-the-art methods quantitatively and qualitatively.

With the advent of cloud computing, renewable energy is integrated into data center power supply systems increasingly. The power statistics collection may not be available due to the instability of renewable energy, which results in incomplete data. The incomplete energy data will significantly disturb the management of data centers. We further propose a filling algorithm based on complete tolerance class. The algorithm expands the traditional tolerance relation, and fills the missing values of the energy data, which ensures the data integrity. By taking good advantage of in-Memory Computing, We further parallelize and optimize our algorithm using Spark. The experiment results demonstrate that our algorithm outperforms other general filling algorithms in terms of filling accuracy. The proposed algorithm also shows good performance as the missing rate rises up.

Capillary nonperfusion (CNP), which is one of characteristic features in diabetic retinopathy patients, is an important judgement of the appearance of retinal diseases. This paper presents a novel, fast and hybrid method to detect CNP. This model combines region-based active contour model(ACM), Graph-Cut and Fuzzy Possibilistic C-Means(FPCM). We modify ACM with a new energy function to reduce suspicious pixels, and employ Graph-Cut to replace the level set method to get a faster arithmetic speed, which also make our method insensitive to the initial contour. And FPCM is applied to average grey levels over different clusters instead of the grey levels of the pixel in order to alleviate the influence caused by noise and capillaries. Experimental results demonstrate that the proposed method outperforms other ACM methods in detecting CNP with better accuracy and sensitivity.

Maintaining the balance is always a challenge issue for bipedal humanoid robots in dealing with various locomotive tasks to serve human society, especially when the real environment the robot worked within exhibits to be very complex. Unlike plenty of previous successful approaches on humanoid balancing or push recovery, in this research, the Dynamical Movement Primitives (DMP) is employed to model several typical bio-inspired strategies. As humanoid balancing or push recovery could be regarded as a problem of how a robot to get back to its ongoing behavior when a break happens on account of external force or uneven terrain etc., the DMP model becomes an alternative ideal choice due to its promising nature of attractor. Meanwhile, the DMP composed of a set of differential equations provides a compact parameterized representation in modelling a motion strategy, and thus leads to a strategy model that is suitable to be fulfilled with machine learning techniques. In this research, the learning process for those bio-inspired strategies modeled with DMP are completed by adopting the stochastic policy gradient reinforcement learning and imitation learning separately. Furthermore, with Gaussian Process(GP) regression, push recovery strategies are generalized taking the advantages of the invariance properties of the DMP model. As a consequence, an online adaptive push recovery control strategy is finally achieved. Experimental results on both simulated robot and a real bipedal humanoid robot PKU-HR5 demonstrate the presented approach is effective and promising.

The main purpose of this paper was to explore the function mechanism between dynamic capabilities and corporate vitality. After concepts introduction of corporate vitality and dynamic capabilities, we offered an integrated theoretical three-dimension model of dynamic capabilities based on analysis of capability attributes and other scholars' studies on dimensions. The three-dimension model consisted of response capacity, organizational learning capacity and organizational innovation capacity. Then we explored the function mechanism between dynamic capabilities and corporate vitality from the three dimensions respectively. After doing that, we put forward a function mechanism model and paths for improving corporate vitality based on dynamic capabilities view. Finally, the function mechanism model was verified by a case study.

A unified pilot protection system is proposed. By virtue of optical fiber communication system, unified pilot protection system can deliver multi-information such as current and voltage sample value, all kinds of digital information. Applying these information, the protection system can not only accomplish many criteria in one system to enhance the performances of pilot protection, also can achieve some other auxiliary functions such as reclosure cooperation, line parameters calculation and so on. Based on the idea of unified pilot protection, this paper studies the solutions of using fault direction information to overcome the problem that differential protection is easily influenced by TA property differences and TA saturation, and gives a detailed analysis of the effect of TA saturation on the performances of current differential component and positive sequence fault direction component. The results show that TA saturation gives limited influence on fault direction component, so it can detect fault direction reliably and provide auxiliary information for current differential protection to distinguish internal fault from external one, accordingly, the paper proposes the corresponding criterion. The results of dynamic simulation experiments verify the feasibility and effectiveness of the proposed scheme. This work is supported by National Natural Science Foundation of China (NO.50807032).

In academic circles, there are some controversial issues at present, such as difference and relation between exergy analysis and exergy transfer, discipline belongingness of exergy transfer and so on. Taking theoretical research in depth on exergy transfer into consideration, it is necessary to clarify these problems. This paper starts with illustrating the scientific definition of energy, and then discusses double property of quantity and quality of energy. The first and second laws of thermodynamics reveal the change law of quantity and quality of energy in energy conversion and transfer process, namely energy quantity conservation and energy quality transmutation, while enthalpy, entropy and exergy are thermodynamic parameters used to describe these laws. Thermodynamics theory is applied to further demonstrate that exergy is a parameter to characterize energy quality. In a word, exergy transfer and energy transfer constitute the complete system of energy research.

The signification of the exergy transfer phenomenological equation for multi-potential fields is introduced, and the concept of field influence factor of exergy transfer resistance is defined. Based on the analysis of the field structure for the thermal driving process, the exergy transfer analysis target is advanced, including the driving work(exergy), driving power(exergy flux), driving resistance(exergy resistance) and the field influence factor of exergy transfer resistance. In the end the aforementioned is illustrated with the hot water driving process. The numerical simulation result can reasonably illuminate the mechanism of some technical measures, such as high temperature decreases viscosity, high pressure gradient increases driving force, high pressure and low water rate decreases driving resistance. Keywords-thermal driving; exergy; exergy analysis; exergy transfer analysis; multi-field cooperation Thermal recovery method such as steam driving, hot water flooding, in situ combustion, is the main means of enhancing heavy oil recovery efficiency. From the point of improving oil displacement effect, the essential description of thermal driving process and its mechanism has special value and practical significance. Oil displacement mechanism which has obtained universal identity at present could be named as 'multiple oil displacement mechanism', it is the comprehensive function of multiple mechanisms including the viscosity reduction of heated crude oil, elastic energy caused by thermal expansion, heat medium displacement, soluble gas driving, miscible drive, gravity drive etc. Its characteristic is that each 'unit' mechanism acts respectively and crossly. Cooperation mechanism among 'multiple' mechanisms is very complex, which results in some key problems related to thermal driving effect getting difficult to solve, such as how to determine the structure composition of 'multiple' mechanism, how to discriminate the thermal driving weight of each 'unit' mechanism and so on, thus it is necesary to explore a mechanism description method which is clear, simple and can felicitously reflect the thermal driving essence. Exergy transfer is the subject to study exergy transferring and transforming rule. It is a crossly frontier subject of thermodynamics and transmission science which was born in the middle of 80's and gradually developed in 90's. To the kind of complex material transmission process such as thermal driving effected simultaneously by various energy forms, it is possible for this theory to transform multi-target into single

Thermodynamic essence of reservoir driving process is studied, which is a typical exergy transfer problem that uses multi-potential fields as driving source to achieve the purpose of oil displacement. The discrimination of reservoir driving mode is derived on the basis of the establishment of driving potential field exergy transfer, which includes driving work (exergy), driving power (exergy flow density), the driving work rate and the driving power rate. Finally with the two-dimensional axial symmetry of hot water flooding process as an example, the main driving mode is distinguished.

At present, most color transfer algorithms are based on single source image, failing to meet reference color demand of object image. This paper proposes a multi-source color transfer algorithm based on multi-labeled decision tree. At first, we define an image as a multi-labeled set since it contains multiple objects. Then every source image is divided into several sub-images to form the training samples for the decision tree. Through extracting color and texture features from sub-image, training dataset is formed and the corresponding label set is obtained by amended K-mean clustering. After that, a multi-labeled decision tree is constructed using SCC_SP. Finally, color transfer is performed on object image based on its predicted label set by the built tree. Experiment results demonstrate that the proposed algorithm works well on multi-source color transfer, making up the shortage of color transfer based on single source image.

In this paper, we propose a novel distributed weighting fault-tolerant algorithm for event region detection in wireless sensor networks. Considering the fault-tolerant range of ldquoneighborhoodpsilas neighborhoodrdquo, we first employ neighborhood nodes to exchange information with its surrounding nodes and estimate the status value of neighborhood nodes, and then weight and synthesize the estimating status value of neighborhood nodes via exploiting weighting fault-tolerant method, thus achieving the fault detection of central nodes. Simulation results show that more than 90 percent of faults can be detected and corrected using this algorithm, even when as many as 20 percent of the nodes are faulty.

Considering a "target" image with complex scene, a novel method for regionally transferring colors from a set of "source" color images to the complex "target" image is presented. An improved EM (expectation-maximization) algorithm is first adopted to approximately model the regional color distribution of the target image by using Gaussian mixture model (GMM); trained by this target GMM, color pixels in each source image are then assigned into source regions corresponding to the target ones, and a color selection mechanism is proposed to choose the best- matched source colors for re-coloring each target region. A new image quality metric, colorfulness, is adopted for objective performance comparisons with previous methods. Experiments show that the proposed method can generate convincing results with well performance.

Diode-pumped Nd:YAG lasers are promising sources of the new optical frequency standards at 532 nm that were recommended recently in the 9th CCDM. Molecular iodine has several strong absorption lines at the second harmonic of Nd:YAG laser near 532 nm, thus providing an absolute reference for frequency stabilization. In general, the 532 nm laser sources used in iodine-stabilized systems are commercial monolithic lasers, which are compact, efficient, and reliable, but are also expensive and complicated. Recently, we have frequency stabilized a self-developed single-frequency Nd:YAG-KTP laser to the Doppler-broadened (instead of Doppler-free) lines of iodine. That was the first time to stabilize a 532 nm laser to an absolute reference in China. Due to its concise structure, low cost, and the ability to resist more disturbance, the laser system is attractive in some certain industrial applications that require an absolute frequency reference but not very high stability.