Search

Existing learned-based image compression methods have shown impressive performance. However, they rely mostly on the consistent distribution between training and test images, which reduces the robustness of the training model. In this paper, we propose a novel compression method called sparse flow adversarial model (SFAM). SFAM employs a deep generative framework to learn a reversible and stable mapping between image distributions, thus it can work in varied scenes for robust compression. The mapping explores the sparsity of the image by combining linear and nonlinear transformations, rather than extracting the features of a particular dataset as is the case with other learning-based methods. Moreover, a sparse adversarial map is introduced into SFAM, to constrain the SFAM to generate sparser features for efficient compression. Extensive experiments are performed on different datasets, in which the effectiveness and robustness of the proposed method are verified. Meanwhile, SFAM is trained only once and it can work well on three different datasets, which also prove the robustness of the proposed SFAM.

The issue of suicide rate has become increasingly rigorous and has received extensive attention in the contemporary society. In order to explore the pattern of suicide rate variation and intrinsic incentive for suicide, the research was conducted to analyze the data involved multifaceted factors since the suicide is rarely caused by only one factor. Establishing statistical models like “Generalized Additive Model” provides a conceptual framework for the study and is instrumental in predicting the suicide rate in 2017. Based on the past data from year 1985 to 2016, the prediction of suicide rate in 2017 is presented specifically. Multiple invisible results are revealed including the negative linear relationship between GDP and suicide rate, the positive correlation with age and the uneven distribution worldwide. These results presented with functions and graphs show how financial, cultural or regional factors might affect the suicide rate in different areas. The prediction are still affected by factors difficult to estimate, such as socio-economic relations. However, the results are sufficient to help understand the suicide issue by analyzing it's origin and distribution.

Distribution network usually refers to the rated voltage of 110KV and below, and the function of direct power supply to users in the power system. The reliability of distribution network power supply directly affects the safety and economy of electricity use. In recent years, with the development of the national economy, the power load has greatly increased. In some areas, the construction of distribution networks cannot keep up with the increase in power load, resulting in power outages and frequent power failures in some regions. At the same time, users are increasingly demanding power quality. Whether the dispatcher or the user is eager to quantify the safety and economic analysis of the distribution network, the accuracy and level of the decision making of the intelligent distribution network dispatcher are improved. Studying the short-circuit fault and single-phase-to-ground fault fast fault location scheme is of great significance to improving the reliability of the distribution network operation. Self-healing is one of the main functions for the safe and reliable operation of smart grids. Reliable detection is the basis for realizing the self-healing function. The probability of a single-phase earth fault in a distribution network is very high because I mostly use it for neutral point grounding in medium-voltage distribution networks and use arc-suppression coils for grounding. When single-phase earthed, it is difficult to find point failures. At present, the small current ground fault line selection technology has been basically mature, and the small current ground fault location technology has achieved good results in the field of transient signals. Based on the transient characteristics of transient components, the single-phase ground fault location technology of power distribution network and the transient characteristics of single-phase grounded neutral distribution network are studied. At the same time, single-phase ground fault detection and cross-section based on parameter identification and current abrupt change are explained. Positioning method, implementation technology of single-phase earth fault feature detection function of distribution automation terminal, and architecture, interface mode and implementation method of single-phase positioning application software for distribution network.