Unidade de Investigação - CIIC - Computer Science and Communication Research Centre
URI permanente desta comunidade:
Navegar
Percorrer Unidade de Investigação - CIIC - Computer Science and Communication Research Centre por Domínios Científicos e Tecnológicos (FOS) "Ciências Naturais::Matemáticas"
A mostrar 1 - 10 de 15
Resultados por página
Opções de ordenação
- Automatic Transcription of Polyphonic Piano Music Using Genetic Algorithms, Adaptive Spectral Envelope Modeling, and Dynamic Noise Level EstimationPublication . Reis, Gustavo; Fernandez de Vega, Francisco; Ferreira, AníbalThis paper presents a new method for multiple fundamental frequency (F0) estimation on piano recordings. We propose a framework based on a genetic algorithm in order to analyze the overlapping overtones and search for the most likely F0 combination. The search process is aided by adaptive spectral envelope modeling and dynamic noise level estimation: while the noise is dynamically estimated, the spectral envelope of previously recorded piano samples (internal database) is adapted in order to best match the piano played on the input signals and aid the search process for the most likely combination of F0s. For comparison, several state-of-the-art algorithms were run across various musical pieces played by different pianos and then compared using three different metrics. The proposed algorithm ranked first place on Hybrid Decay/Sustain Score metric, which has better correlation with the human hearing perception and ranked second place on both onset-only and onset–offset metrics. A previous genetic algorithm approach is also included in the comparison to show how the proposed system brings significant improvements on both quality of the results and computing time.
- Customized crowds and active learning to improve classificationPublication . Costa, Joana; Silva, Catarina; Antunes, Mário; Ribeiro, BernardeteTraditional classification algorithms can be limited in their performance when a specific user is targeted. User preferences, e.g. in recommendation systems, constitute a challenge for learning algorithms. Additionally, in recent years user’s interaction through crowdsourcing has drawn significant interest, although its use in learning settings is still underused. In this work we focus on an active strategy that uses crowd-based non-expert information to appropriately tackle the problem of capturing the drift between user preferences in a recommendation system. The proposed method combines two main ideas: to apply active strategies for adaptation to each user; to implement crowdsourcing to avoid excessive user feedback. A similitude technique is put forward to optimize the choice of the more appropriate similitude-wise crowd, under the guidance of basic user feedback. The proposed active learning framework allows non-experts classification performed by crowds to be used to define the user profile, mitigating the labeling effort normally requested to the user. The framework is designed to be generic and suitable to be applied to different scenarios, whilst customizable for each specific user. A case study on humor classification scenario is used to demonstrate experimentally that the approach can improve baseline active results.
- Discrete Compound Tests and Dorfman’s Methodology in the Presence of MisclassificationPublication . Santos, Rui; Martins, João Paulo; Felgueiras, MiguelCompound tests can be used to save resources for classification or estimation purposes in clinical trials and quality control. Nevertheless, the methodologies that are usually applied are restricted to qualitative group tests. Moreover, when quantitative compound tests are applied the problem is to ascertain whether the amount of some substance of any individual in the group is greater or lower than a prefixed threshold. An overview of the applications of the discrete compound tests highlights the advantages (to save resources) and disadvantages (higher probability of misclassification), and suggests criteria to assess the suitability of applying Dorfman’s methodology.
- Evolution of Artificial Terrains for Video Games Based on AccessibilityPublication . Frade, Miguel; Vega, Francisco Fernandez de; Cotta, CarlosDiverse methods have been developed to generate terrains under constraints to control terrain features, but most of them use strict restrictions. However, there are situations were more flexible restrictions are sufficient, such as ensuring that terrains have enough accessible area, which is an important trait for video games. The Genetic Terrain Program technique, based on genetic programming, was used to automatically evolve Terrain Programs (TPs - which are able to generate terrains procedurally) for the desired accessibility parameters. Results showed that the accessibility parameters have negligible influence on the evolutionary system and that the terminal set has a major role on the terrain look. TPs produced this way are already being used on Chapas video game.
- Exploring Pareto scale mixturesPublication . Felgueiras, Miguel; Santos, RuiPareto scale mixtures can be used to obtain distributions with heavier tails. An explanation of this model properties is provided, together with a discussion about the parameters estimation. Finally, a real data application is presented, consisting in the larger earthquakes seismic moment modeling.
- Forensic Analysis of Tampered Digital PhotosPublication . Ferreira, Sara; Antunes, Mário; Correia, Manuel E.Deepfake in multimedia content is being increasingly used in a plethora of cybercrimes, namely those related to digital kidnap, and ransomware. Criminal investigation has been challenged in detecting manipulated multimedia material, by applying machine learning techniques to distinguish between fake and genuine photos and videos. This paper aims to present a Support Vector Machines (SVM) based method to detect tampered photos. The method was implemented in Python and integrated as a new module in the widely used digital forensics application Autopsy. The method processes a set of features resulting from the application of a Discrete Fourier Transform (DFT) in each photo. The experiments were made in a new and large dataset of classified photos containing both legitimate and manipulated photos, and composed of objects and faces. The results obtained were promising and reveal the appropriateness of using this method embedded in Autopsy, to help in criminal investigation activities and digital forensics.
- Hybrid population-based incremental learning to assign terminals to concentratorsPublication . Bernardino, Eugénia Moreira; Bernardino, Anabela Moreira; Sánchez-Pérez, Juan Manuel; Gómez-Pulido, Juan Antonio; Vega-Rodríguez, Miguel AngelIn the last decade, we have seen a significant growth in communication networks. In centralised communication networks, a central computer serves several terminals or workstations. In large networks, some concentrators are used to increase the network efficiency. A collection of terminals is connected to a concentrator and each concentrator is connected to the central computer. In this paper we propose a Hybrid Population-based Incremental Learning (HPBIL) to assign terminals to concentrators. We use this algorithm to determine the minimum cost to form a network by connecting a given collection of terminals to a given collection of concentrators. We show that HPBIL is able to achieve good solutions, improving the results obtained by previous approaches.
- A Hybrid Scatter Search algorithm to assign terminals to concentratorsPublication . Bernardino, Eugénia Moreira; Bernardino, Anabela Moreira; Sánchez-Pérez, Juan Manuel; Gómez-Pulido, Juan Antonio; Vega-Rodríguez, Miguel AngelThe last few years have seen a significant growth in communication networks. With the growth of data traffic, network operators seek network-engineering tools to extract the maximum benefits out of the existing infrastructure. This has suggested a number of new optimisation problems, most of them in the field of combinatorial optimisation. We address here the Terminal Assignment problem. The main objective is to assign a collection of terminals to a collection of concentrators. In this paper, we propose a Hybrid Scatter Search (HSS) algorithm to assign terminals to concentrators. Coupled with the Scatter Search algorithm we use a Tabu Search algorithm to locate the global minimum. We show that HSS is able to achieve feasible solutions to Terminal Assignment instances, improving the results obtained by previous approaches.
- Indicator-based multi objective evolutionary algorithms and an application in filament winding processPublication . Yevseyeva, Iryna; de Melo, Francisco Queirós; Grácio, José; Basto-Fernandes, VitorThis work presents recent developments on multi objective evolutionary algorithms, so-called set-based evolutionary algorithms. These techniques are shown to approximate a Pareto front of efficient solutions taking into account both quality of the approximation and its diversity, both important in the design of these methods. Set-based evolutionary algorithms outperform their predecessors on a variety of benchmark problems and are suggested as tools to be used for solving complex mechanical engineering problems, such as filament winding process discussed in this work.
- Pareto Scale MixturesPublication . Felgueiras, Miguel MartinsPareto scale mixtures are very effective for modelling heavy-tailed data. A new class of models is described, generalizing commonly used slash distributions. Mixture properties and possible applications are discussed.