STRUCTURE AND REQUIREMENTS FOR A UNIVERSAL TRAINING DEVICE FOR UNMANNED AIRCRAFT COMPLEXES

The problems of personnel training for the operation of unmanned aerial vehicles have been updated. The differences between the skills and abilities of personnel for manned and unmanned aircraft are highlighted. The widespread types of flight simulators are considered, their features are given. The basic schemes of application of linear and nonlinear mathematical models of formation (formalization) of knowledge, abilities and skills of the personnel taking into account absence / presence of updating of the information in the course of training are substantiated.The basic modes of operation of USL, the universal simulator for training are specified. Emphasis is placed on the fact that USL is a partial case of an automated training management system (ACS), and a block diagram of training with it is given. The tasks that can be solved on a universal simulator and the requirements for software and hardware, automated workplace, digital sound, general software, peripherals, data exchange, digital video, user interface and webbased automated instructions are discussed in detail. The functions and structure of the universal simulator for training unmanned aircraft complex are presented.

Multi-View Active Learning for Video Recommendation

On many video websites, the recommendation is implemented as a prediction problem of video-user pairs, where the videos are represented by text features extracted from the metadata. However, the metadata is manually annotated by users and is usually missing for online videos. To train an effective recommender system with lower annotation cost, we propose an active learning approach to fully exploit the visual view of videos, while querying as few annotations as possible from the text view. On one hand, a joint model is proposed to learn the mapping from visual view to text view by simultaneously aligning the two views and minimizing the classification loss. On the other hand, a novel strategy based on prediction inconsistency and watching frequency is proposed to actively select the most important videos for metadata querying. Experiments on both classification datasets and real video recommendation tasks validate that the proposed approach can significantly reduce the annotation cost.

“There Was No Green Tick”: Discovering the Functions of a Widget in a Joint Problem-Solving Activity and the Consequences for the Participants’ Discovering Process

In recent years, tangible user interfaces (TUI) have gained in popularity in educational contexts, among others to implement problem-solving and discovery learning science activities. In the context of an interdisciplinary and cross-institutional collaboration, we conducted a multimodal EMCA-based video user study involving a TUI-mediated bicycle mechanics simulation. This article focusses on the discovering work of a group of three students with regard to a particular tangible object (a red button), designed to support participants engagement with the underlying physics aspects and its consequences with regard to their engagement with the targeted mechanics aspects.