The theme and variation music form is a hierarchical structure in music. It has a theme segment at the beginning, followed by a series of variation segments imitating the theme segment. Hence, the primary features of the theme and variation form are repetition and variation at different levels. However, due to the lack of available datasets, the theme and variation form analysis method has not been explored much. Therefore, in this work, we curate and contribute a dataset named Performance of Theme and Variation Form (PTV) and propose a theme and variation form segmentation framework to analyze the theme and variation form. Experiment results show that our method achieves an F1 score of 92.8% on our dataset with some constraints. In addition, we conduct analysis to support and encourage future studies of the theme and variation form.

Abstract As an essential part of music, main melody is the cornerstone of music information retrieval. In the MIR’s sub-field of main melody extraction, the mainstream methods assume that the main melody is unique. However, the assumption cannot be established, especially for music with multiple main melodies such as symphony or music with many harmonies. Hence, the conventional methods ignore some main melodies in the music. To solve this problem, we propose a deep learning-based Multiple Main Melodies Generator (Multi-MMLG) framework that can automatically predict potential main melodies from a MIDI file. This framework consists of two stages: (1) main melody classification using a proposed MIDIXLNet model and (2) conditional prediction using a modified MuseBERT model. Experiment results suggest that the proposed MIDIXLNet model increases the accuracy of main melody classification from 89.62 to 97.37%. In addition, this model requires fewer parameters (71.8 million) than the previous state-of-art approaches. We also conduct ablation experiments on the Multi-MMLG framework. In the best-case scenario, predicting meaningful multiple main melodies for the music are achieved.

Classifying music pieces based on their instrument sounds is pivotal for analysis and application purposes. Given its importance, techniques using machine learning have been proposed to classify violin and viola music pieces. The violin and viola are two different instruments with three overlapping strings of the same notes, and it is challenging for ordinary people or even musicians to distinguish the sound produced by these instruments. However, the classification of musical instrument pieces was barely performed by prior research. To solve this problem, we propose a technique using descriptive statistics to reliably distinguish between violin and viola music pieces. Likewise, a similar technique on the basis of histogram is introduced alongside the main descriptive statistics approach. These approaches are derived based on the nature of the instruments’ strings and the range of their pieces. We also solve the problem in the current literature which divide the audio into segments for processing instead of managing the whole song. Thereby, we compile a dataset of recordings that comprises of violin and viola solo pieces from the Baroque, Classical, Romantic, and Modern eras. Experiment results suggest that our approach achieves high accuracy on solo pieces as compared to other methods with 0.97 accuracy on Baroque pieces.

Summary As a crowdsourcing map platform, OpenStreetMap (OSM) relies on public contributions to enhance its dataset where the contributors can create, modify or remove features from the maps or share their trajectory trips in the repository. The majority of the data provided in a crowdsourcing platform are manually created and reviewed to suit real‐world conditions, hence human perception is the key indicator to consider the correctness of the data. One of the data that is provided by crowdsourcing platform is public trajectory. Public trajectory data contains details of historical trips obtained from contributors' GPS logger devices that are embedded in mobile devices, wearable devices, satnavs, or vehicle GPS trackers to record the user's trajectory path. While public trajectory data can be used as an alternate data source for human movement analysis, this crowdsourced dataset is also prone to noise and inaccuracy which makes the preprocessing step an important phase prior of any processing step. In this article, we discuss the characteristics and the most common noise from crowdsourcing GPS trajectories and utilize a non‐map‐matching approach convex hull‐based reduction method to minimize spike noise, followed by granularity reduction to reduce the number of trajectory points while maintaining the nature of the trajectories.

Trajectory is the path that is formed because of the moving object positioning history. Based on where it happened, trajectory can be formed in indoor or outdoor environment. While outdoor trajectory can be reconstructed by GPS technology, this technology is not sufficient to be used in indoor environment. However, there are some embedded sensors in mobile devices that can be utilized to track and reconstruct trajectory in indoor environment without utilizing GPS technology. In this paper, we propose the method to track and reconstruct trajectory for indoor environment using embedded mobiles sensors. Our experiment shows that these sensors are capable in tracking and reconstructing indoor trajectories without using any GPS technologies.

Navigation is capable of showing the position and direction at the specified location. In this paper, we proposed an indoor navigation based on the user location to its desired destination. Indoors navigation first needs to determine the route that can be accessed inside the building. This guidance is made by cultivating the indoor trajectory route that utilizes the sensor technologies in mobile devices. The sensors to be used are pedometer and magnetometer. The experiment shows that the guidance gave the instructions to navigate from the start to the destination by following compass direction and footstep to take. The guidance instruction affected by three factors, there is user height, step calculation, and sensor threshold value. These three factors have different effects but interconnected with the application system and that which causes differences in the accuracy values obtained in the experiment.

Emergency is an event that can threaten and disrupt a someone's life. Emergency situation occur suddenly with unexpected place and time. Different emergency situations may need different emergency unit type, either polices, ambulances or fire fighters. Most countries have implemented single number to handle emergency situation where the users will get the adequate emergency units as they need. To identify the coverage of each emergency unit for each emergency type, a multi-layered network Voronoi diagram is used. Then, finding nearest emergency unit and alternate units from a certain location can be determined by using Kth-Nearest Neighbor algorithm applied in predefined multi-layered network Voronoi diagram..

In Reverse Nearest Neighbour (RNN) Query, every single object in the space has a certain region where all objects inside this region will think of the query object as their nearest neighbour. Other objects that are outside the region will not consider the query object as their nearest neighbour. In many cases, we might encounter a situation where we want to find this kind of region for several objects altogether, instead of a single object. Current RNN region approach cannot be used for this problem. Therefore we propose Group Reverse kNN as a solution, which we will find a specific region based on multiple query objects. So any objects located inside this region will always consider all of the query objects as the nearest compare to the non-query objects. Our experiments demonstrate the performance efficiency of the proposed Group Reverse kNN algorithm.

Reverse Nearest Neighbour queries is known for the heavyweight algorithm that makes it difficult to be implemented in mobile devices due to high computations needed to verify the objects. Since the rapid development of mobile devices' hardware and also the new lightweight approach in solving the Reverse Nearest Neighbour problem, this problem can be solved in mobile devices. In this paper, we implemented the Contact Zone algorithm to create bichromatic reverse nearest neighbour region for a specific query point on mobile devices. The model is developed in a closed wireless network and various types of mobile devices with different hardware specifications are used. Our experiments show that RNN queries by region can be solved in mobile devices and different mobile CPUs do not give significant performance in processing the queries.

SUMMARY Common reverse nearest neighbor queries in spatial database run in an inefficient way because they need to check a query result with almost every nearest neighbor. This wastes many time and resources, making this approach unsuitable for mobile computation. Instead of using the neighbors as candidates for the query result, a region approach can be used to answer the query. By using this approach, any objects located in the region will be considered candidate results for the query. To reduce the cost of creating the region, we introduce the concept of a contact zone, a method that can identify the right region generator points without having to process the whole points in the space, hence make reverse nearest neighbor queries by region possible to be run in mobile devices. Copyright © 2013 John Wiley & Sons, Ltd.

The increasing use of mobile communications has raised many issues of decision support and resource allocation. A crucial problem is how to solve queries of Reverse Nearest Neighbor (RNN). An RNN query returns all objects that consider the query object as their nearest neighbor. Existing methods mostly rely on a centralized base station. However, mobile P2P systems offer many benefits, including self-organization, fault-tolerance and load-balancing. In this study, we propose two P2P algorithms focusing on bichromatic RNN queries, in which mobile query objects and static objects of interest are of two different categories, based on a boundary polygon around the mobile query object. The Exhaustive Search Algorithm makes use of all information from the peers to aim at high accuracy rate while the Optimized Search Algorithm reduces the number of queried peers. The algorithms are evaluated in MiXiM simulation framework with a real dataset. The results show the practical feasibility of the P2P approach in solving bichromatic RNN queries for mobile networks.