Search

Image saliency detection is to segment the most important areas in the image. Solving the problem of image saliency detection usually involves knowledge in computer vision, neuroscience, cognitive psychology and other fields. In recent years, as deep learning has made great achievements in the field of computer vision, the application of deep learning has also played a good role in image saliency detection. Therefore, algorithms based on deep convolutional neural networks have become solutions to image saliency The most effective method of detection. For researchers, improving the computational efficiency of neural network-based saliency detection algorithms generally starts from two perspectives. One is to tailor the network structure and combine traditional feature extraction methods for processing. The other is to use a lighter network to solve the saliency detection problem. Based on these two points of thinking, this paper proposes two efficient and accurate neural network-based saliency detection algorithms. In recent years, with the rapid development of multimedia and Internet technologies, a huge amount of picture information is generated on blogs, social networking or shopping platforms every day. Such a lot of information not only enriches people’s lives, but also provides efficient and accurate network management platforms. The management of these image information brings difficulties. Therefore, how to understand and process these image information more intelligently and efficiently has become a hot topic for many image processing and computer vision researchers. Among them, the saliency detection technology plays a key role in solving the problem of intelligent understanding and processing of images. To put it simply, saliency detection is a technology to automatically calculate or detect the most important areas in an image, and its processing results provide a basis for understanding and processing the image content. Saliency detection is a basic problem in computer vision, neuroscience and visual perception. The algorithm detects and extracts the most interesting or significant areas in the image.

To address the problems in the traditional speech signal noise elimination methods, such as the residual noise, poor real-time performance and narrow applications a new method is proposed to eliminate network voice noise based on deep learning of conditional generation adversarial network. In terms of the perceptual evaluation of speech quality (PESQ) and shorttime objective intelligibility measure (STOI) functions used as the loss function in the neural network, which were used as the loss function in the neural network, the flexibility of the whole network was optimized, and the training process of the model simplified. The experimental results indicate that, under the noisy environment, especially in a restaurant, the proposed noise reduction scheme improves the STOI score by 26.23% and PESQ score by 17.18%, respectively, compared with the traditional Wiener noise reduction algorithm. Therefore, the sound sensor’s noise reduction scheme through our approach has achieved a remarkable noise reduction effect, more useful information transmission, and stronger practicability.

Based on the power supply management needs of power supply companies, this paper adopts a big data-driven power outage data collection method to build a scientific customer outage time key indicator around the power grid comprehensive customer service system, establish a customer outage time indicator responsibility system, and establish a customer outage The process monitoring platform has formed a data-driven customer power outage analysis system to effectively handle customer power outages.

Monopole to ground operation mode of ultra high-voltage DC transmission system will result in DC magnetic biasing phenomenon to transformers in surrounding AC substation. Under monopole to ground operation mode of ±800kV "Qi-Shao" ultra high-voltage DC transmission project, DC magnetic biasing hazard of transformers in AC power network in sending terminal—Jiuquan region was tested. It's found that after capacitive DC blocking device of 750kV transformer neutral point was put into action, transformer vibration and noise still gradually increased with ground current at ground electrode, so ideal governance effect was not reached. Regional DC current distribution after series capacitance at transformer neutral point in 750kV Dunhuang substation was analyzed in this paper. It's found that DC current could still flow through autotransformer winding so as to make the transformer be still influenced by DC magnetic bias. Meanwhile, based on analysis of secondary-side waveforms at neutral point of 750kV autotransformer, it's found that transformer magnetizing current went through serious distortion with current increase at ground electrode after action of DC blocking device. Harmonic analysis was conducted for distorted waveforms and change features of secondary harmonics were used, which provided a convenient and effective means of monitoring DC magnetic biasing hazard of 750kV autotransformer.

We report on simultaneous magneto-optical trapping of fermionic 40K and bosonic 87Rb atoms. This trap is the first step towards quantum degenerate fermi gas 40K. Laser lights for the two-species magneto-optical trap (MOT) are generated from diode lasers and tapered amplifier. The enriched 40K dispenser is utilized in the experimental setup. We obtain up to 107~108 40K and 108~109 87Rb atoms respectively in the steady-state single-species MOT.

Validity of adiabatic assumption is discussed within the constrained Hartree-Fock theory for self-conjugate nucleus Kr-72. It is shown that the adiabatic assumption does not provide a correct description for the nature of nucleus when a, configuration change is involved. The excited Hartree-Fock states and the continuously-connected constrained Hartree-Fock states are given for the first time by applying the configuration dictated constrained Hartree-Fock theory with Gogny force. The importance of self-consistency between the mean-field and the single particle wave functions is emphasized even when a small number of nucleons are involved in the configuration change.

Ground-state properties of C, O, and Ne isotopes are described in the framework of Hartree–Fock–Bogoliubov theory with density-dependent finite-range Gogny interaction D1S. We include all the contributions to the Hartree–Fock and pairing field arising from Gogny and Coulomb interaction as well as the center of mass correction in the numerical calculations. These ground-state properties of C, O, and Ne isotopes are compared with available experimental results, Hartree–Fock plus BCS, shell model and relativistic Hartree–Bogoliubov calculations. The agreement between experiments and our theoretical results is pretty well. The predicted drip-line is dependent strongly on the model and effective interaction due to their sensitivity to various theoretical details. The calculations predict no evidence for halo structure predicted for C, O, and Ne isotopes in a previous RHB study.

Energy level statistics of a system consisting of six particles interacting by delta force in a two-j model coupled with a deformed core is studied in particle-rotor model. For single-j shell (i(13/2)) and two-j shell (g(7/2) + d(5/2)) the exact energies for our statistical analysis are obtained from a full diagonalization of the Hamiltonian, while in two-j case (i(13/2) + g(9/2)) the configuration truncation is used. The nearest-neighbor distribution of energy levels and spectral rigidity are studied as the function of spin. The results of single-j shell are compared with those in two-j case. It is showed that the system becomes more regular when single-j space (i(13/2)) is replaced by two-j shell (g(7/2) + d(5/2)) although the basis size of the configuration space is unchanged. The degree of chaoticity of the system, however, changes slightly when configuration space is enlarged by extending single-j shell (i(13/2)) to two-j shell (i(13/2) + g(9/2)).

The chaotic properties for six particles interacting by a monopole pairing force in a two-j shell model coupled with a deformed core are studied in the frame of particle-rotor model. The nearest-neighbor distribution of energy levels and spectral rigidity in the two-j shell are compared with those in the single-j case. The results show that the system is more regular in the two-j model than that in the single-j case.

Magnetic radiation phenomena appear inevitably in the magnetic-resonance wireless power transfer (MR-WPT) system, and regarding this problem the magnetic-shielding scheme is applied to improve the electromagnetic performance in engineering. In this study, the shielding effectiveness of a two-coil MR-WPT system for different material shields is analyzed in theory using Moser's formula and Schelkunoff's formula. On this basis a candidate magnetic-shielding scheme with a double-layer structure is determined, which has better shielding effectiveness and coils coupling coefficient. Finally, some finite element simulation results validate the correctness of the theoretical analysis, and the shielding effectiveness with the double-layer shield in maximum is 30 dB larger than the one with the single-layer case.

It is shown that the flux motion in high-Tc superconductors can be well consistently analysed by a unified materials equation. The different models of U(J) so far suggested can be shown as specific forms or limits of this unified materials equation. The scaling law of flux pinning derived from this equation agrees with recent experimental observations on high-Jc NdBa2Cu3O7-δ single crystal. The amplitude dependence of ac susceptibility derived from this equation is compared with several experimental results of high-Tc superconductors and a general power law is found.

Structural properties and thermodynamic properties of amorphous SiO2(a-SiO2) and amorphous SiO2 with an oxygen defect center (a-SiO2-ODC) are investigated by using interatomic pair potentials. Amorphous configurations of SiO2 glass are constructed by quenching using the molecular-dynamics method and subsequent relaxation using the first-principles method. The numerical result demonstrates that the calculated short-range structural parameters coincide well with the previous experimental and theoretical results reported. The thermodynamic properties including specific heat, Debye temperature, vibrational entropy, and so on are analyzed by calculating the phonon density. Furthermore, mechanical properties of both structures are obtained.