Search

Most of the existing image tampering detection methods are only for a certain tampering method, and there is a problem that the detection accuracy of the tampered area boundary is not high. In this regard, this paper proposed a U-shaped network based on dual-stream encoder-decoder architecture for image tampering. Firstly, the method used the skip connection between the encoder and the decoder to fuse the low-level and high-level features in the tampered image, and used the atrous convolution and CBAM attention mechanism to fuse the features output by the encoder, so that the network had better localization performance for tampered regions of different scales; Secondly, in order to improve the network’s detection accuracy of the boundary of the tampered area, this algorithm used the image morphological method to make a tampered boundary dataset; Finally, it used multiple loss functions to optimize the performance of the network simultaneously, namely it used cross-entropy and root-mean-square loss functions to measure the tampered region loss and tampered boundary loss of the prediction graph respectively. The experimental results on 4 public datasets: CASIA, Columbia, NIST16, Coverage show that the method proposed in this paper can effectively detect the tampering area of the fake image with splicing and copy-paste tampering methods, and output pixel-level tampering region location map. Compared with other mainstream tampering detection methods, the proposed method achieves the highest AUC values on the CASIA and Columbia dataset and the highest F1 values on the Columbia dataset. [ABSTRACT FROM AUTHOR]

This document presents in detail the mathematical formulation of the second renewable energy auction in Colombia, used for the allocation of long-term energy contracts, which was carried out through a double-sided auction, defined in Resolutions 4-0590 and 4-0591 of 2019 from the UPME. The mixed integer linear programming model developed in this paper solves an optimization problem that seeks the combination of offers that maximizes the benefit of the consumer, subject to operational and economic constraints proposed by regulatory mechanisms. The effectiveness of the algorithm developed in Python was tested by validating it with real data, obtaining the same results published by UPME. [ABSTRACT FROM AUTHOR]

The concept of algorithmic political communication has arisen through the joint development of propaganda and communication theories, as well as the findings of computer science. This is a field that generates results that amplify and extend to political communication, such as microsegmentation or automated diffusion of content; likewise, it produces some adverse effects that hinder citizen participation in the cybersphere. This paper presents a review of works published in three Latin American countries (Colombia, Ecuador, and Mexico) with artificial intelligence, communication, and democracy as their constituent elements. The data come from a total of 206 documents, including reports from supranational organisms and associations, and publications in scientific journals indexed in databases such as Google Scholar and Scopus between 2011 and 2021. The selection criterion is based on the total or partial appearance of certain keywords, plus filters by relevance and impact factors. This systematic review is structured based along four axes that explain democracy according to Landman et al. (2009): electoral participation, participation of civil society, media integrity, and impartial administration. The main conclusions indicate that many of the practices in the studied countries are still at an incipient or experimental stage, with algorithmic political communication being used especially during electoral campaigns. Similarly, an increasing resistance from civil society to the influence of social networks is starting to be detected. [ABSTRACT FROM AUTHOR]

This paper proposes a numerical method for the calibration of dynamic stochastic general equilibrium (DSGE) models. The method consists in the use of a hybrid algorithm, first to find the steady state, and then to minimize an objective function defined by the researcher according to the purposes of the calibration. The proposed method consists in the use of the Simulated Annealing algorithm followed by traditional optimization routines. The virtues of the algorithm are analyzed through Monte Carlo simulations, using a closed-economy model that has a steady state with no analytical solution. The results obtained in this exercise show that the proposed algorithm generates more precise results using less computational resources than traditional alternatives. Finally we present the calibration of a model for the Colombian economy, consisting of 179 equations and adjusted through 50 parameters to replicate 50 ratios. The maximum percent deviation of the steady-state ratios of the model with respect to the corresponding values in the Colombian economy is 7.9%, and in 29 of all the 50 cases this deviation is less than or equal to 1%. [ABSTRACT FROM AUTHOR]

In this paper, we analyse the bipartite Colombian firms-products network, throughout a period of five years, from 2010 to 2014. Our analysis depicts a strongly modular system, with several groups of firms specializing in the export of specific categories of products. These clusters have been detected by running the bipartite variant of the traditional modularity maximization, revealing a bi-modular structure. Interestingly, this finding is refined by applying a recently proposed algorithm for projecting bipartite networks on the layer of interest and, then, running the Louvain algorithm on the resulting monopartite representations. Important structural differences emerge upon comparing the Colombian firms-products network with the World Trade Web, in particular, the bipartite representation of the latter is not characterized by a similar block-structure, as the modularity maximization fails in revealing (bipartite) nodes clusters. This points out that economic systems behave differently at different scales: while countries tend to diversify their production—potentially exporting a large number of different products—firms specialize in exporting (substantially very limited) baskets of basically homogeneous products. [ABSTRACT FROM AUTHOR]