O codificador aritmético binário adaptativo ao contexto adotado (CABAC – Context-based Adaptive Binary Arithmetic Coding) pelo padrão H.264/AVC a partir de perfil Main é o estado-da-arte em termos de eficiência de taxa de bits. Entretanto, o CABAC ocupa 9.6% do tempo total de processamento e seu throughput é limitado pelas dependências de dados no nível de bit (LIN, 2010). Logo, atingir os requisitos de desempenho em tempo real nos níveis mais altos do padrão H.264/AVC se torna uma tarefa árdua em software, sendo necesário então, a aceleração do CABAC através de implementações em hardware. As arquiteturas de hardware encontradas na literatura para o CABAC focam no Codificador Aritmético Binário (BAE - Binary Arithmetic Encoder) enquanto que a Binarização e Modelagem de Contextos (BCM – Binarization and Context Modeling) fica em segundo plano ou nem é apresentada. O BCM e o BAE juntos constituem o CABAC. Esta dissertação descreve detalhadamente o conjunto de algoritmos que compõem o BCM do padrão H.264/AVC. Em seguida, o projeto de uma arquitetura de hardware específica para o BCM é apresentada. A solução proposta é descrita em VHDL e os resultados de síntese mostram que a arquitetura alcança desempenho suficiente, em FPGA e ASIC, para processar vídeos no nível 5 do padrão H.264/AVC. A arquitetura proposta é 13,3% mais rápida e igualmente eficiente em área que os melhores trabalhos relacionados nestes quesitos. / Context-based Adaptive Binary Arithmetic Coding (CABAC) adopted in the H.264/AVC main profile is the state-of-art in terms of bit-rate efficiency. However, CABAC takes 9.6% of the total encoding time and its throughput is limited by bit-level data dependency (LIN, 2010). Moreover, meeting real-time requirement for a pure software CABAC encoder is difficult at the highest levels of the H.264/AVC standard. Hence, speeding up the CABAC by hardware implementation is required. The CABAC hardware architectures found in the literature focus on the Binary Arithmetic Encoder (BAE), while the Binarization and Context Modeling (BCM) is a secondary issue or even absent in the literature. Integrated, the BCM and the BAE constitute the CABAC. This dissertation presents the set of algorithms that describe the BCM of the H.264/AVC standard. Then, a novel hardware architecture design for the BCM is presented. The proposed design is described in VHDL and the synthesis results show that the proposed architecture reaches sufficiently high performance in FPGA and ASIC to process videos in real-time at the level 5 of H.264/AVC standard. The proposed design is 13.3% faster than the best works in these items, while being equally efficient in area.

Software systems in domains like Smart Cities, the Internet of Things or autonomous cars are coined by a high degree of distribution across several independent computing devices and the requirement to be able to adjust themselves to varying situations in their operational environment. Self-adaptive software systems are a natural choice to implement such context-dependent software systems. A multitude of approaches already implement self-adaptive systems and some consider even distribution aspects.Yet, none of the existing solutions supports the coordination of adaptation operations spanning multiple independent nodes, which is necessary to ensure a consistent adaptation even in presence of network errors or node failures. In this paper, we tackle this challenge to execute adaptations in distributed self-adaptive software systems in a coordinated manner. We present a protocol that enables the self-adaptive software system to execute correlated adaptations on multiple nodes in a transactional manner ensuring an atomic and consistent transition of the distributed system from its source to the desired target configuration. The protocol is validated to be free of deadlocks for any given adaptation at any point in time using a model-checking approach. The performance of our approach is investigated in experiments that emulate the protocol's execution on real devices for different sizes of distributed applications and adaptation scenarios.

Made available in DSpace on 2015-05-14T12:36:29Z (GMT). No. of bitstreams: 1 arquivototal.pdf: 2909809 bytes, checksum: 27b0cc11fa58c3080c9bf2558be74b71 (MD5) Previous issue date: 2011-08-22 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / Due to particular interactivy mode performed by TV services, engines are requeried that allow to retrieve informations beyond those provided directly. In this sense, context-aware applications use relevant informations to provide support in tasks execution. To develop these kind of applications presents challenges in context capture from heterogeneous sources, as sensors; by representation more adjusted to perform context-aware behavior; and to enable infer knowledges. This dissertation proposes an approach for content personalization of Interactive TV Applications by context handling. To this end, a context modelling was achieved to describe the user information and sports content information semantic in an integrated way and services were developed whose features provides support to context usage. / Graças ao modo peculiar de interatividade realizada por aplicativos na TV, são necessários mecanismos que possibilitem recuperar informações além daquelas fornecidas diretamente. Neste sentido, aplicações sensíveis ao contexto utilizam informações consideradas relevantes para fornecer suporte à realização de tarefas. Desenvolver aplicações desta natureza apresenta desafios quanto a captura de dados a partir de diferentes fontes, tais como sensores; quanto a representação mais adequada para realizar comportamento sensível ao contexto; e capacitar a inferência de conhecimentos. Este trabalho propõe uma abordagem para personalização do conteúdo de aplicações de TV Digital Interativa através da manipulação de informações de contexto. Para tanto, foi realizada a especificação de um modelo contextual que descreve semântica de informações do usuário e de conteúdo esportivo de maneira integrada, e foram desenvolvidos serviços cujas funcionalidades oferecem suporte ao uso de contexto.

Multimodal Emotion Recognition (MER) has gained increasing attention due to its exceptional performance. In this thesis, we evaluate feature-level fusion, decision-level fusion, and two proposed hierarchical fusion methods for MER systems using a dialogue-based dataset. The first hierarchical approach integrates abstract features across different temporal levels by employing RNN-based and transformer-based context modeling techniques to capture nearby and global context respectively. The second hierarchical strategy incorporates shared information between modalities by facilitating modality interactions through attention mechanisms. Results reveal that RNN-based hierarchical fusion surpasses the baseline by 2%, while transformer-based context modeling and modality interaction methods improve accuracy by 0.5% and 0.6%, respectively. These findings underscore the significance of capturing meaningful emotional cues in nearby context and emotional invariants in dialogue MER systems. We also emphasize the crucial role of text modality. Overall, our research highlights the potential of hierarchical fusion approaches for enhancing MER system performance, presenting systematic strategies supported by empirical evidence. / Multimodal Emotion Recognition (MER) har fått ökad uppmärksamhet på grund av dess exceptionella prestanda. I denna avhandling utvärderar vi feature-level fusion, decision-level fusion och två föreslagna hierarkiska fusion-metoder för MER-system med hjälp av en dialogbaserad dataset. Den första hierarkiska metoden integrerar abstrakta funktioner över olika tidsnivåer genom att använda RNN-baserade och transformer-baserade tekniker för kontextmodellering för att fånga närliggande och globala kontexter, respektive. Den andra hierarkiska strategin innefattar delad information mellan modaliteter genom att underlätta modalitetsinteraktioner genom uppmärksamhetsmekanismer. Resultaten visar att RNN-baserad hierarkisk fusion överträffar baslinjen med 2%, medan transformer-baserad kontextmodellering och modellering av modalitetsinteraktion ökar noggrannheten med 0.5% respektive 0.6%. Dessa resultat understryker betydelsen av att fånga meningsfulla känslomässiga ledtrådar i närliggande sammanhang och emotionella invarianter i dialog MER-system. Vi betonar också den avgörande rollen som textmodalitet spelar. Övergripande betonar vår forskning potentialen för hierarkiska fusion-metoder för att förbättra prestandan i MER-system, genom att presentera systematiska strategier som stöds av empirisk evidens.