The representation of the physical world is an issue that concerns the scientific community studying computer vision, more and more. Recently, research has focused on modern techniques and methods of photogrammetry and stereoscopy with the aim of reconstructing three-dimensional realistic models with high accuracy and metric information in a short time. In order to obtain data at a relatively low cost, various tools have been developed, such as depth cameras. RGB-D cameras are novel sensing systems that capture RGB images along with per-pixel depth information. This survey aims to describe RGB-D camera technology. We discuss the hardware and data acquisition process, in both static and dynamic environments. Depth map sensing techniques are described, focusing on their features, pros, cons, and limitations; emerging challenges and open issues to investigate are analyzed; and some countermeasures are described. In addition, the advantages, disadvantages, and limitations of RGB-D cameras in all aspects are also described critically. This survey will be useful for researchers who want to acquire, process, and analyze the data collected.

Keystroke dynamics are used to authenticate users, to reveal some of their inherent or acquired characteristics and to assess their mental and physical states. The most common features utilized are the time intervals that the keys remain pressed and the time intervals that are required to use two consecutive keys. This paper examines which of these features are the most important and how utilization of these features can lead to better classification results. To achieve this, an existing dataset consisting of 387 logfiles is used, five classifiers are exploited and users are classified by gender and age. The results, while demonstrating the application of these two characteristics jointly on classifiers with high accuracy, answer the question of which keystroke dynamics features are more appropriate for classification with common classifiers.

Over the past decade, keystroke-based pattern recognition techniques, as a forensic tool for behavioral biometrics, have gained increasing attention. Although a number of machine learning-based approaches have been proposed, they are limited in terms of their capability to recognize and profile a set of an individual's characteristics. In addition, up to today, their focus was primarily gender and age, which seem to be more appropriate for commercial applications (such as developing commercial software), leaving out from research other characteristics, such as the educational level. Educational level is an acquired user characteristic, which can improve targeted advertising, as well as provide valuable information in a digital forensic investigation, when it is known. In this context, this paper proposes a novel machine learning model, the randomized radial basis function network, which recognizes and profiles the educational level of an individual who stands behind the keyboard. The performance of the proposed model is evaluated by using the empirical data obtained by recording volunteers' keystrokes during their daily usage of a computer. Its performance is also compared with other well-referenced machine learning models using our keystroke dynamic datasets. Although the proposed model achieves high accuracy in educational level prediction of an unknown user, it suffers from high computational cost. For this reason, we examine ways to reduce the time that is needed to build our model, including the use of a novel data condensation method, and discuss the tradeoff between an accurate and a fast prediction. To the best of our knowledge, this is the first model in the literature that predicts the educational level of an individual based on the keystroke dynamics information only.

In this paper an incident response approach is proposed for handling detections of authentication failures in systems that employ dynamic biometric authentication and more specifically keystroke user recognition. The main component of the approach is a multi layer perceptron focusing on the age classification of a user. Empirical findings show that the classifier can detect the age of the subject with a probability that is far from the uniform random distribution, making the proposed method suitable for providing supporting yet circumstantial evidence during e-discovery.

Cracks can occur on different surfaces such as buildings, roads, aircrafts, etc. The manual inspection of cracks is time-consuming and prone to human error. Machine vision has been used for decades to detect defects in materials in production lines. However, the detection or segmentation of cracks on a randomly textured surface, such as marble, has not been sufficiently investigated. This work provides an up-to-date systematic and exhaustive study on marble crack segmentation with color images based on deep learning (DL) techniques. The authors conducted a performance evaluation of 112 DL segmentation models with red–green–blue (RGB) marble slab images using five-fold cross-validation, providing consistent evaluation metrics in terms of Intersection over Union (IoU), precision, recall and F1 score to identify the segmentation challenges related to marble cracks’ physiology. Comparative results reveal the FPN model as the most efficient architecture, scoring 71.35% mean IoU, and SE-ResNet as the most effective feature extraction network family. The results indicate the importance of selecting the appropriate Loss function and backbone network, underline the challenges related to the marble crack segmentation problem, and pose an important step towards the robotic automation of crack segmentation and simultaneous resin application to heal cracks in marble-processing plants.