Cookies
Usamos cookies para melhorar nosso site e a sua experiência. Ao continuar a navegar no site, você aceita a nossa política de cookies. Ver mais
Aceitar Rejeitar
  • Menu
Sobre
Download foto HD

Sobre

Maria Teresa Andrade é Professora Auxiliar na FEUP, no DEEC. Obteve a licenciatura em 1986, o grau de mestre em 1992 e o doutoramento em 2008 em Eng. Electrotécnica e de Computadores pela FEUP. Participa em actividades de investigação no INESC TEC, integrada na área de Sistemas Multimédia do Centro de Telecomunicações e Multimédia. Áreas de interesse: aplicações multimédia sensíveis ao contexto de utilização, em ambientes móveis e heterogéneos; tecnologias semânticas e recomendação de conteúdos; streaming de vídeo 3D e multi-vista; Qualidade de serviço e de experiência em serviços multimédia; televisão e cinema digitais e novos media.

Tópicos
de interesse
Detalhes

Detalhes

011
Publicações

2019

Predictive multi-view content buffering applied to interactive streaming system

Autores
Costa, TS; Andrade, MT; Viana, P;

Publicação
Electronics Letters

Abstract

2019

Adaptation execution cost definition for a multimedia adaptation decision engine using a neural network

Autores
Fernandes, R; Andrade, MT;

Publicação
AIP Conference Proceedings

Abstract
Multimedia content adaptation decision is necessary whenever a multimedia transmission system has multiple adaptations available to adjust the content representation requirements to the present available system resources. The implementation of an adaptation decision module, based on a Markov Decision Process, requires to weight the adaptations, to establish the adaptation plan to deliver the best possible Quality of Experience (QoE) to the user. We present a method, using a feedforward neural network, to determine these costs using two approaches: user and service provider perspectives. © 2019 Author(s).

2018

GymApp: A real time physical activity trainner on wearable devices

Autores
Viana, P; Ferreira, T; Castro, L; Soares, M; Pinto, JP; Andrade, MT; Carvalho, P;

Publicação
Proceedings - 2018 11th International Conference on Human System Interaction, HSI 2018

Abstract
Technological advances are pushing into the mass market innovative wearable devices featuring increasing processing and sensing capacity, non-intrusiveness and ubiquitous use. Sensors built-in those devices, enable acquiring different types of data and by taking advantage of the available processing power, it is possible to run intelligent applications that process the sensed data to offer added-value to the user in multiple domains. Although not new to the modern society, it is unquestionable that the present exercise boom is rapidly spreading across all age groups. However, in a great majority of cases, people perform their physical activity on their own, either due to time or budget constraints and may easily get discouraged if they do not see results or perform exercises inadequately. This paper presents an application, running on a wearable device, aiming at operating as a personal trainer that validates a set of proposed exercises in a sports session. The developed solution uses inertial sensors of an Android Wear smartwatch and, based on a set of pattern recognition algorithms, detects the rate of success in the execution of a planned workout. The fact that all processing can be executed on the device is a differentiator factor to other existing solutions. © 2018 IEEE.

2018

Multimedia Content Classification Metrics for Content Adaptation

Autores
Fernandes, R; Andrade, MT;

Publicação
U.Porto Journal of Engineering

Abstract
Multimedia content consumption is very popular nowadays. However, not every content can be consumed in its original format: the combination of content, transport and access networks, consumption device and usage environment characteristics may all pose restrictions to that purpose. One way to provide the best possible quality to the user is to adapt the content according to these restrictions as well as user preferences. This adaptation stage can be best executed if knowledge about the content is known a-priori. In order to provide this knowledge we classify the content based on metrics to define its temporal and spatial complexity. The temporal complexity classification is based on the Motion Vectors of the predictive encoded frames and on the difference between frames. The spatial complexity classification is based on different implementations of an edge detection algorithm and an image activity measure.

2018

ML datasets as synthetic cognitive experience records

Autores
Castro, H; Andrade, MT;

Publicação
International Journal of Computer Information Systems and Industrial Management Applications

Abstract
Machine Learning (ML), presently the major research area within Artificial Intelligence, aims at developing tools that can learn, approximately on their own, from data. ML tools learn, through a training phase, to perform some association between some input data and some output evaluation of it. When the input data is audio or visual media (i.e. akin to sensory information) and the output corresponds to some interpretation of it, the process may be described as Synthetic Cognition (SC). Presently ML (or SC) research is heterogeneous, comprising a broad set of disconnected initiatives which develop no systematic efforts for cooperation or integration of their achievements, and no standards exist to facilitate that. The training datasets (base sensory data and targeted interpretation), which are very labour intensive to produce, are also built employing ad-hoc structures and (metadata) formats, have very narrow expressive objectives and thus enable no true interoperability or standardisation. Our work contributes to overcome this fragility by putting forward: a specification for a standard ML dataset repository, describing how it internally stores the different components of datasets, and how it interfaces with external services; and a tool for the comprehensive structuring of ML datasets, defining them as Synthetic Cognitive Experience (SCE) records, which interweave the base audio-visual sensory data with multilevel interpretative information. A standardised structure to express the different components of the datasets and their interrelations will promote re-usability, resulting on the availability of a very large pool of datasets for a myriad of application domains. Our work thus contributes to: the universal interpretability and reusability of ML datasets; greatly easing the acquisition and sharing of training and testing datasets within the ML research community; facilitating the comparison of results from different ML tools; accelerating the overall research process. © MIR Labs.

Teses
supervisionadas

2019

Dashcam Video Streaming and Storage

Autor
José Miguel Matos Lopes da Costa

Instituição
UP-FEUP

2019

knowledge and content-based strategies for 3d vídeo content adptation decision

Autor
Rui Vítor Pires Fernandes

Instituição
UP-FEUP

2019

Programação Visual para a Definição de um Estúdio Televisivo Baseado na Tecnologia IP

Autor
Margarida Xavier Viterbo

Instituição
UP-FEUP

2019

Data Analytics Mechanisms for Sport's Related Data Retrieval and Multimedia Augmentation

Autor
Tiago André Pérola Filipe

Instituição
UP-FEUP

2019

Real-time Video Fusion for Surveillance Applications

Autor
André da Cunha Trancoso Ranito

Instituição
UP-FEUP