Cookies
O website necessita de alguns cookies e outros recursos semelhantes para funcionar. Caso o permita, o INESC TEC irá utilizar cookies para recolher dados sobre as suas visitas, contribuindo, assim, para estatísticas agregadas que permitem melhorar o nosso serviço. Ver mais
Aceitar Rejeitar
  • Menu
Sobre

Sobre

Jaime S. Cardoso, licenciado em Engenharia e Eletrotécnica e de Computadores em 1999, Mestre em Engenharia Matemática em 2005 e doutorado em Visão Computacional em 2006, todos pela Universidade do Porto. Professor Associado com agregação na Faculdade de Engenharia da Universidade do Porto (FEUP) e Investigador Sénior em 'Information Processing and Pattern Recognition' no Centro de Telecomunicações e Multimédia do INESC TEC.

A sua investigação assenta em três grandes domínios: visão computacional, "machine learning" e sistemas de suporte à decisão. A investigação em processamento de imagem e vídeo tem abordado a área de biometria, imagem médica e "video tracking" para aplicações de vigilância e desportos. O trabalho em "machine learning" foca-se na adaptação de sistemas de aprendizagem às condições desafiantes de informação visual. A ênfase dos sistemas de suporte à decisão tem sido dirigida a aplicações médicas, sempre ancoradas com a análise automática de informação visual.

É co-autor de mais de 150 artigos, dos quais mais de 50 em jornais internacionais, com mais de 6500 citações (google scholar). Foi investigador principal em 6 projectos de I&D e participou em 14 projectos de I&D, incluindo 5 projectos europeus e um contrato directo com a BBC do Reino Unido.

Tópicos
de interesse
Detalhes

Detalhes

021
Publicações

2022

Lesion Volume Quantification Using Two Convolutional Neural Networks in MRIs of Multiple Sclerosis Patients

Autores
de Oliveira, M; Piacenti Silva, M; da Rocha, FCG; Santos, JM; Cardoso, JD; Lisboa, PN;

Publicação
DIAGNOSTICS

Abstract
Background: Multiple sclerosis (MS) is a neurologic disease of the central nervous system which affects almost three million people worldwide. MS is characterized by a demyelination process that leads to brain lesions, allowing these affected areas to be visualized with magnetic resonance imaging (MRI). Deep learning techniques, especially computational algorithms based on convolutional neural networks (CNNs), have become a frequently used algorithm that performs feature self-learning and enables segmentation of structures in the image useful for quantitative analysis of MRIs, including quantitative analysis of MS. To obtain quantitative information about lesion volume, it is important to perform proper image preprocessing and accurate segmentation. Therefore, we propose a method for volumetric quantification of lesions on MRIs of MS patients using automatic segmentation of the brain and lesions by two CNNs. Methods: We used CNNs at two different moments: the first to perform brain extraction, and the second for lesion segmentation. This study includes four independent MRI datasets: one for training the brain segmentation models, two for training the lesion segmentation model, and one for testing. Results: The proposed brain detection architecture using binary cross-entropy as the loss function achieved a 0.9786 Dice coefficient, 0.9969 accuracy, 0.9851 precision, 0.9851 sensitivity, and 0.9985 specificity. In the second proposed framework for brain lesion segmentation, we obtained a 0.8893 Dice coefficient, 0.9996 accuracy, 0.9376 precision, 0.8609 sensitivity, and 0.9999 specificity. After quantifying the lesion volume of all patients from the test group using our proposed method, we obtained a mean value of 17,582 mm3 . Conclusions: We concluded that the proposed algorithm achieved accurate lesion detection and segmentation with reproducibility corresponding to state-of-the-art software tools and manual segmentation. We believe that this quantification method can add value to treatment monitoring and routine clinical evaluation of MS patients. © 2022 by the authors. Licensee MDPI, Basel, Switzerland.

2022

Tackling unsupervised multi-source domain adaptation with optimism and consistency

Autores
Pernes, D; Cardoso, JS;

Publicação
EXPERT SYSTEMS WITH APPLICATIONS

Abstract
It has been known for a while that the problem of multi-source domain adaptation can be regarded as a single source domain adaptation task where the source domain corresponds to a mixture of the original source domains. Nonetheless, how to adjust the mixture distribution weights remains an open question. Moreover, most existing work on this topic focuses only on minimizing the error on the source domains and achieving domain-invariant representations, which is insufficient to ensure low error on the target domain. In this work, we present a novel framework that addresses both problems and beats the current state of the art by using a mildly optimistic objective function and consistency regularization on the target samples. © 2022 Elsevier Ltd

2022

Streamlining Action Recognition in Autonomous Shared Vehicles with an Audiovisual Cascade Strategy

Autores
Pinto, JR; Carvalho, P; Pinto, C; Sousa, A; Capozzi, L; Cardoso, JS;

Publicação
PROCEEDINGS OF THE 17TH INTERNATIONAL JOINT CONFERENCE ON COMPUTER VISION, IMAGING AND COMPUTER GRAPHICS THEORY AND APPLICATIONS (VISAPP), VOL 5

Abstract

2022

Myope Models - Are face presentation attack detection models short-sighted?

Autores
Neto, PC; Sequeira, AF; Cardoso, JS;

Publicação
2022 IEEE/CVF WINTER CONFERENCE ON APPLICATIONS OF COMPUTER VISION WORKSHOPS (WACVW 2022)

Abstract

2022

Quasi-Unimodal Distributions for Ordinal Classification

Autores
Albuquerque, T; Cruz, R; Cardoso, JS;

Publicação
MATHEMATICS

Abstract
Ordinal classification tasks are present in a large number of different domains. However, common losses for deep neural networks, such as cross-entropy, do not properly weight the relative ordering between classes. For that reason, many losses have been proposed in the literature, which model the output probabilities as following a unimodal distribution. This manuscript reviews many of these losses on three different datasets and suggests a potential improvement that focuses the unimodal constraint on the neighborhood around the true class, allowing for a more flexible distribution, aptly called quasi-unimodal loss. For this purpose, two constraints are proposed: A first constraint concerns the relative order of the top-three probabilities, and a second constraint ensures that the remaining output probabilities are not higher than the top three. Therefore, gradient descent focuses on improving the decision boundary around the true class in detriment to the more distant classes. The proposed loss is found to be competitive in several cases.

Teses
supervisionadas

2021

Make My Heartbeat: Generation and Interlead Conversion of ECG Signals

Autor
Sofia Cardoso Beco

Instituição
UP-FEUP

2021

Label-efficient learning of LiDAR-based perception models for autonomous driving

Autor
Bernardo Magina Madureira Palha de Araújo

Instituição
UP-FEUP

2021

Protecting IOT devices against cyberattacks using deep learning to classify radio-frequency emissions of code running on the processor

Autor
José Pedro Passos Leocádio

Instituição
UP-FEUP

2021

Fast prototyping of advanced sensing devices using three-dimensional direct writing with femtosecond laser

Autor
Vítor Alexandre Oliveira Amorim

Instituição
UP-FCUP

2021

Scene understanding from 3D point clouds and RGB images for autonomous driving

Autor
Rolando de Sousa Chichorro Avides Moreira

Instituição
UP-FEUP