The global community has recognized an increasing amount of pollutants entering oceans and other water bodies as a severe environmental, economic, and social issue. In addition to prevention, one of the key measures in addressing marine pollution is the cleanup of debris already present in marine environments. Deployment of machine learning (ML) and deep learning (DL) techniques can automate marine waste removal, making the cleanup process more efficient. This study examines the performance of six well-known deep convolutional neural networks (CNNs), namely VGG19, InceptionV3, ResNet50, Inception-ResNetV2, DenseNet121, and MobileNetV2, utilized as feature extractors according to three different extraction schemes for the identification and classification of underwater marine debris. We compare the performance of a neural network (NN) classifier trained on top of deep CNN feature extractors when the feature extractor is (1) fixed; (2) fine-tuned on the given task; (3) fixed during the first phase of training and fine-tuned afterward. In general, fine-tuning resulted in better-performing models but is much more computationally expensive. The overall best NN performance showed the fine-tuned Inception-ResNetV2 feature extractor with an accuracy of $91.40\%$ and F1-score $92.08\%$, followed by fine-tuned InceptionV3 extractor. Furthermore, we analyze conventional ML classifiers’ performance when trained on features extracted with deep CNNs. Finally, we show that replacing NN with a conventional ML classifier, such as support vector machine (SVM) or logistic regression (LR), can further enhance the classification performance on new data.

Unmanned Aircraft Systems (UASs) have been recognized as an important resource in search-and-rescue (SAR) missions and, as such, have been used by the Croatian Mountain Search and Rescue (CMRS) service for over seven years. The UAS scans and photographs the terrain. The high-resolution images are afterwards analyzed by SAR members to detect missing persons or to find some usable trace. It is a drawn out, tiresome process prone to human error. To facilitate and speed up mission image processing and increase detection accuracy, we have developed several image-processing algorithms. The latest are convolutional neural network (CNN)-based. CNNs were trained on a specially developed image database, named HERIDAL. Although these algorithms achieve excellent recall, the efficiency of the algorithm in actual SAR missions and its comparison with expert detection must be investigated. A series of mission simulations are planned and recorded for this purpose. They are processed and labelled by a developed algorithm. A web application was developed by which experts analyzed raw and processed mission images. The algorithm achieved better recall compared to an expert, but the experts achieved better accuracy when they analyzed images that were already processed and labelled.

This paper defines an advanced methodology for modeling applications based on Data Mining methods that represents a logical framework for development of Data Mining applications. Methodology suggested here for Data Mining modeling process has been applied and tested through Data Mining applications for predicting Prepaid users churn in the telecom industry. The main emphasis of this paper is defining of a successful model for prediction of potential Prepaid churners, in which the most important part is to identify the very set of input variables that are high enough to make the prediction model precise and reliable. Several models have been created and compared on the basis of different Data Mining methods and algorithms (neural networks, decision trees, logistic regression). For the modeling examples we used WEKA analysis tool.

Search and rescue (SAR) missions comprise search for, and provision of aid to people who are in distress or imminent danger. Providing the best possible input for the planners and search teams, up-to-date information about the terrain is of essential importance because every additional hour needed to search a person decreases probability of success. Therefore, availability of aerial images and updated terrain maps as a basis for planning and monitoring SAR missions in real-time is very important for rescuers. In this paper, we present a system for transmission of high-resolution images from an unmanned aerial vehicle (UAV) to the ground station (GS). We define and calculate data rate and transmission distance requirements between the UAV and GS in a mission scenario. Five tests were designed and carried out to confirm the viability of the proposed system architecture and modules. Test results present throughput measurements for various UAV and GS distances, antenna heights and UAV antenna yaw angles. Experimental results from the series of conducted outdoor tests show that the proposed solution using two pMDDL2450 datalinks at 2.4 GHz and a directional antenna on the receiving side can be used for a real-time transmission of high-resolution images acquired with a camera on a UAV. Achieved throughput at a UAV-GS distance of 5 km was 1.4 MB/s (11.2 Mbps). The limitations and possible improvements of the proposed system as well as future work are also discussed.

High-energy hadronic collisions generate environments characterized by temperatures above 100 MeV (refs. ^1,2 ), about 100,000 times hotter than the centre of the Sun. At present, it is therefore unclear how light (anti)nuclei with mass number A of a few units, such as the deuteron, ³ He or ⁴ He, each bound by only a few MeV, can emerge from these collisions ^3,4 . Here, the ALICE Collaboration reports that deuteron–pion momentum correlations in proton–proton (pp) collisions provide model-independent evidence that about 90% of the observed (anti)deuterons are produced in nuclear reactions ⁵ following the decay of short-lived resonances, such as the Δ(1232). These findings, obtained by the ALICE Collaboration at the Large Hadron Collider, resolve a gap in our understanding of nucleosynthesis in ultrarelativistic hadronic collisions. Apart from offering insights on how (anti)nuclei are formed in hadronic collisions, the results can be used in the modelling of the production of light and heavy nuclei in cosmic rays ⁶ and dark-matter decays ^7,8 .