Data leakage is a problem that companies and organizations face every day around the world. Mainly the data leak caused by the internal threat posed by authorized personnel to manipulate confidential information. The main objective of this work is to survey the literature to detect the existing techniques to protect against data leakage and to identify the methods used to address the insider threat. For this, a literature review of scientific databases was carried out in the period from 2011 to 2022, which resulted in 42 relevant papers. It was obtained that from 2017 to date, 60% of the studies found are concentrated and that 90% come from conferences and publications in journals. Significant advances were detected in protection systems against data leakage with the incorporation of new techniques and technologies, such as machine learning, blockchain, and digital rights management policies. In 40% of the relevant studies, significant interest was shown in avoiding internal threats. The most used techniques in the analyzed DLP tools were encryption and machine learning.
This paper presents a new adaptive downsampling technique called elastic downsampling, which enables high compression rates while preserving the image quality. Adaptive downsampling techniques are based on the idea that image tiles can use different sampling rates depending on the amount of information conveyed by each block. However, current approaches suffer from blocking effects and artifacts that hinder the user experience. To bridge this gap, elastic downsampling relies on a Perceptual Relevance analysis that assigns sampling rates to the corners of blocks. The novel metric used for this analysis is based on the luminance fluctuations of an image region. This allows a gradual transition of the sampling rate within tiles, both horizontally and vertically. As a result, the block artifacts are removed and fine details are preserved. Experimental results (using the Kodak and USC Miscelanea image datasets) show a PSNR improvement of up to 15 dB and a superior SSIM (Structural Similarity) when compared with other techniques. More importantly, the algorithms involved are computationally cheap, so it is feasible to implement them in low-cost devices. The proposed technique has been successfully implemented using graphics processors (GPU) and low-power embedded systems (Raspberry Pi) as target platforms.
Video streaming over IP networks presents several challenges for remote drone piloting. To achieve a high Quality of Experience, minimal latency is mandatory. However, wireless links usually impose dynamic changes to Quality of Service conditions. Moreover, bandwidth limitations can increase both the final perceived latency and packet loss during video streaming. These circumstances require an architecture capable of estimating network performance and applying corrective actions in a timely manner to optimize application-level quality. In this paper, we present such an architecture, and discuss the results of its application in video streaming for remote drone piloting. Our proposal offers a framework with low coupling between its functional blocks and high adaptability to dynamic scenarios. Accordingly, we aim to pave the way for reactive applications that leverage edge-computing elements and adapt to network conditions.
The growing demand for real-time video streaming in power-constrained embedded systems, such as drone navigation and remote surveillance, requires encoding solutions that prioritize low latency. In these applications, even small delays in video transmission can impair the operator’s ability to react in time, leading to instability in closed-loop control systems. To mitigate this, encoding must be lightweight and designed so that streaming can start as soon as possible, ideally even while frames are still being processed, thereby ensuring continuous and responsive operation. This paper presents the design of a hardware implementation of the Logarithmic Hop Encoding (LHE) algorithm on a Field-Programmable Gate Array (FPGA). The proposed architecture is deeply pipelined and parallelized to achieve sub-frame latency. It employs adaptive compression by dividing frames into regions of interest and uses a quantized differential system to minimize data transmission. Our design achieves an encoding latency of between 1.87 ms and 2.1 ms with a power consumption of only 2.7 W when implemented on an FPGA clocked at 150 MHz. Compared to a parallel GPU implementation of the same algorithm, this represents a 6.6-fold reduction in latency at approximately half the power consumption. These results show that FPGA-based LHE is a highly effective solution for low-latency, real-time video applications and establish a robust foundation for its deployment in embedded systems.
2 September 2025 in International Journal of Medical Informatics
Insider threats pose a critical risk in healthcare environments, where Hospital Information Systems (HIS) manage sensitive patients data. Authorized users may intentionally or accidentally compromise data confidentiality, integrity, and availability. This study assessed information security practices from the perspective of healthcare professionals in Spanish medical centers. A descriptive, analytical, cross-sectional study was conducted using a survey administered to 41 healthcare professionals with access to confidential data. The survey covered access control, encryption at rest and in transit, communication channels, and data usage control. Descriptive statistics, Chi-square tests, and Cramér's V were applied to identify significant associations. K-means clustering and Silhouette coefficient were used to define user profiles. Principal Component Analysis (PCA) was used to visualize behavior patterns. A Random Forest model identified the most relevant predictive variables. Critical security gaps were detected, 31.7 % reported no control over data usage. Only 29.3 % encrypted data at rest and 36.6 % during transmission. Over 40 % used personal email or messaging apps to share sensitive data, and 97.6 % relied solely on passwords for authentication. These practices are inadequate to mitigate insider threats. There is an urgent need to strengthen insider data protection. Security strategies should be tailored to user risk profiles. Measures must include strong authentication, full encryption, and stricter control of data transmission to reduce exposure to insider threats (intentionally or unintentionally) in healthcare settings. Additionally, there is a need to promote continuous cybersecurity training.
13 January 2026 in ADCAIJ: Advances in Distributed Computing and Artificial Intelligence Journal
Nowadays, deep learning models are quickly increasing in complexity and size, making the training or inference phases unsuitable for most office computers or even high-performance computing servers. Examples of such models include, but are not limited to, large language models and image diffusion models, where the number of parameters to adjust can reach hundreds of billions. Although deep learning models are extremely useful tools, the need for powerful computers to train and use them might compromise the obtained results and also reduce accessibility for many researchers. In this work, we propose a general method to reduce the training time and computing resources (central processing unit and memory usage) by the division of a large-size model into several smaller submodels that are easier to handle but do not affect the final performance. This division depends on the original model and architecture, and hence we propose specific strategies for regression and classification problems. The main result of this work is the development of a public Python library called Skynnet to offer this alternative to deep learning computations.