Abstract
The Malware Information Sharing Platform (MISP) enables the sharing of cyberthreat information within a community, company or organisation. However, this platform presents limitations if its information is deemed as classified or shared only for a given period of time. This implies that this information should to be handled only in encrypted form. One solution is to use MISP with searchable encryption techniques to impose greater control over the sharing of information. In this paper, we propose a controlled information sharing functionality that features a synchronisation procedure that enables classified data exchange between MISP instances, based on policies and ensuring the required confidentiality and integrity of the shared data. Sequence charts are presented validating the configuration, the data synchronisation, and the data searching between multiple entities.

Abstract
The actual learning process in a school, college or university should take full advantage of the digital transformation. Computers, mobile phones, tablets or other electronic devices can be used in learning environments to improve learning experience and students performance. However, in a university campus, there are some activities where the use of connected devices, might be discouraged or even forbidden. Students should be discouraged to use their own devices in classes where they may become alienated or when their devices may cause any disturbance. Ultimately, their own devices should be forbidden in activities such as closed-book exams. This paper proposes a system architecture to detect or block unwanted wireless signals by students' mobile phones in a classroom. This architecture focuses on specific wireless signals from Wi-Fi and Bluetooth interfaces, and it is based on Software-Defined Radio (SDR) modules and a set of antennas with two configuration modes: detection mode and blocking mode. When in the detection mode, the architecture processes signals from the antennas, detects if there is any signal from Wi-Fi or Bluetooth interfaces and infers a position of the unwanted mobile device. In the blocking mode, the architecture generates noise in the same frequency range of Wi-Fi or Bluetooth interfaces, blocking any possible connection. The proposed architecture is designed to be used by professors to detect or block unwanted wireless signals from student devices when supervising closed-book exams, during specific periods of time. © Daniel Barros, Paulo Barros, Emanuel Lomba, Vítor Ferreira, and Pedro Pinto; licensed under Creative Commons License CC-BY 4.0 Second International Computer Programming Education Conference (ICPEC 2021).

Abstract

Abstract

Abstract
Rhythmic similarity, a fundamental task within Music Information Retrieval, has recently been applied in creative music contexts to retrieve musical audio or guide audio-content transformations. However, there is still very little knowledge of the typical rhythmic similarity values between overlapping musical structures per instrument, genre, and time scales, which we denote as rhythmic compatibility. This research provides the first steps towards the understanding of rhythmic compatibility from the systematic analysis of MedleyDB, a large multi-track musical database composed and performed by artists. We apply computational methods to compare database stems using representative rhythmic similarity metrics - Rhythmic Histogram (RH) and Beat Spectrum (BS) - per genre and instrumental families and to understand whether RH and BS are prone to discriminate genres at different time scales. Our results suggest that 1) rhythmic compatibility values lie between [.002,.354] (RH) and [.1,.881] (BS), 2) RH outperforms BS in discriminating genres, and 3) different time scale in RH and BS impose significant differences in rhythmic compatibility. © 2021 ACM.

Abstract
In this paper, we address a subtopic of the broad domain of audio enhancement, namely musical audio bandwidth extension. We formulate the bandwidth extension problem using deep neural networks, where a band-limited signal is provided as input to the network, with the goal of reconstructing a full-bandwidth output. Our main contribution centers on the impact of the choice of low-pass filter when training and subsequently testing the network. For two different state-of-the-art deep architectures, ResNet and U-Net, we demonstrate that when the training and testing filters are matched, improvements in signal-to-noise ratio (SNR) of up to 7 dB can be obtained. However, when these filters differ, the improvement falls considerably and under some training conditions results in a lower SNR than the band-limited input. To circumvent this apparent overfitting to filter shape, we propose a data augmentation strategy which utilizes multiple low-pass filters during training and leads to improved generalization to unseen filtering conditions at test time.