This work focuses on a cognitive science inspired solution to an event detection problem in a video domain. The thesis raises the question whether video sequences that are taken in highway tunnels can be used to create meaningful data in terms of symbolic representation, and whether these symbolic representations can be used as sequences to be parsed by attribute grammars into abnormal and normal events. The main motivation of the research was to develop a novel algorithm that parses sequences of primitive events created by the image processing algorithms. The domain of the research is video detection and the special application purpose is for highway tunnels, which are critical places for abnormality detection. The method used is attribute grammars to parse the sequences. The symbolic sequences are created from a cascade of image processing algorithms such as / background subtracting, shadow reduction and object tracking. The system parses the sequences and creates alarms if a car stops, moves backwards, changes lanes, or if a person walks into the road or is in the vicinity when a car is moving along the road. These critical situations are detected using Earley&rsquo / s parser, and the system achieves real-time performance while processing the video input. This approach substantially lowers the number of false alarms created by the lower level image processing algorithms by preserving the number of detected events at a maximum. The system also achieves a high compression rate from primitive events while keeping the lost information at minimum. The output of the algorithm is measured against SVM and observed to be performing better in terms of detection and false alarm performance.

The availability of digital media has grown tremendously with the fast-paced ever-growing storage and communication technologies. As a result, today, we are facing a problem in indexing and browsing the huge amounts of multimedia data. This amount of data is impossible to be indexed or browsed by hand so automatic indexing and browsing systems are proposed. Audio Event Detection is a research area which tries to analyse the audio data in a semantic and perceptual manner, to bring a conceptual solution to this problem. In this thesis, a method for detecting several audio events in TV broadcast is proposed. The proposed method includes an audio segmentation stage to detect event boundaries. Broadcast audio is classified into 17 classes. The feature set for each event is obtained by using a feature selection algorithm to select suitable features among a large set of popular descriptors. Support Vector Machines and Gaussian Mixture Models are used as classifiers and the proposed system achieved an average recall rate of 88% for 17 different audio events. Comparing with the results in the literature, the proposed method is promising.

Esta dissertação trata do projeto e avaliação de um novo circuito sensor de corrente com célula de memória dinâmica para a detecção de correntes transientes em circuitos integrados CMOS, provocadas pela incidência de partículas ionizantes. As propostas previamente existentes na literatura são avaliadas e suas deficiências são apontadas. É apresentada a topologia e o modo de funcionamento do novo circuito, juntamente com o detalhamento do projeto das versões destinadas à monitoração dos transistores PMOS e NMOS. É apresentado o layout do circuito final em tecnologia 130 nm, destinado à prototipação pelo programa MOSIS, contendo os sensores, os transistores-alvo, os estágios de saída e os circuitos de proteção contra os efeitos da eletricidade estática necessários. Os resultados obtidos através de simulação mostram que o novo circuito proporciona uma redução na área de silício necessária para a implementação, bem como um menor consumo de corrente quiescente em relação às propostas anteriores. / This dissertation deals with the design and evaluation of a new current sensor circuit with dynamic memory cell intended to detect transient currents caused by incidence of ionizing particles in CMOS integrated circuits. Circuits previously proposed are analyzed and their drawbacks are pointed out. The new circuit topology and working principle is presented, along with the detailed design of the versions intended to monitoring PMOS and NMOS transistors. The final circuit is laid out in a 130 nm technology, intended to be prototyped through the MOSIS program. The complete design contains the sensor circuits, target transistors, output stages and electrostatic discharge protection circuitry. Results obtained by post layout simulation shown that the new circuit provides a reduction on silicon area and a smaller quiescent current consumption compared to previous circuits.

Esta dissertação trata do projeto e avaliação de um novo circuito sensor de corrente com célula de memória dinâmica para a detecção de correntes transientes em circuitos integrados CMOS, provocadas pela incidência de partículas ionizantes. As propostas previamente existentes na literatura são avaliadas e suas deficiências são apontadas. É apresentada a topologia e o modo de funcionamento do novo circuito, juntamente com o detalhamento do projeto das versões destinadas à monitoração dos transistores PMOS e NMOS. É apresentado o layout do circuito final em tecnologia 130 nm, destinado à prototipação pelo programa MOSIS, contendo os sensores, os transistores-alvo, os estágios de saída e os circuitos de proteção contra os efeitos da eletricidade estática necessários. Os resultados obtidos através de simulação mostram que o novo circuito proporciona uma redução na área de silício necessária para a implementação, bem como um menor consumo de corrente quiescente em relação às propostas anteriores. / This dissertation deals with the design and evaluation of a new current sensor circuit with dynamic memory cell intended to detect transient currents caused by incidence of ionizing particles in CMOS integrated circuits. Circuits previously proposed are analyzed and their drawbacks are pointed out. The new circuit topology and working principle is presented, along with the detailed design of the versions intended to monitoring PMOS and NMOS transistors. The final circuit is laid out in a 130 nm technology, intended to be prototyped through the MOSIS program. The complete design contains the sensor circuits, target transistors, output stages and electrostatic discharge protection circuitry. Results obtained by post layout simulation shown that the new circuit provides a reduction on silicon area and a smaller quiescent current consumption compared to previous circuits.

Esta dissertação trata do projeto e avaliação de um novo circuito sensor de corrente com célula de memória dinâmica para a detecção de correntes transientes em circuitos integrados CMOS, provocadas pela incidência de partículas ionizantes. As propostas previamente existentes na literatura são avaliadas e suas deficiências são apontadas. É apresentada a topologia e o modo de funcionamento do novo circuito, juntamente com o detalhamento do projeto das versões destinadas à monitoração dos transistores PMOS e NMOS. É apresentado o layout do circuito final em tecnologia 130 nm, destinado à prototipação pelo programa MOSIS, contendo os sensores, os transistores-alvo, os estágios de saída e os circuitos de proteção contra os efeitos da eletricidade estática necessários. Os resultados obtidos através de simulação mostram que o novo circuito proporciona uma redução na área de silício necessária para a implementação, bem como um menor consumo de corrente quiescente em relação às propostas anteriores. / This dissertation deals with the design and evaluation of a new current sensor circuit with dynamic memory cell intended to detect transient currents caused by incidence of ionizing particles in CMOS integrated circuits. Circuits previously proposed are analyzed and their drawbacks are pointed out. The new circuit topology and working principle is presented, along with the detailed design of the versions intended to monitoring PMOS and NMOS transistors. The final circuit is laid out in a 130 nm technology, intended to be prototyped through the MOSIS program. The complete design contains the sensor circuits, target transistors, output stages and electrostatic discharge protection circuitry. Results obtained by post layout simulation shown that the new circuit provides a reduction on silicon area and a smaller quiescent current consumption compared to previous circuits.

No description available.

This report presents the design and manufacturing process of a bionic signal messagebroker (BSMB), intended to allow communication between implanted electrodes andprosthetic legs designed by Ossur. The BSMB processes and analyses the data intorelevant information to control the bionic device. The intention is to carry out eventdetection in the BSMB, where events in the muscle signal are matched to the events ofthe gait cycle (toe-o, stance, swing).The whole system is designed to detect muscle contraction via sensors implantedin residual muscles and transmit the signals wireless to a control unit that activatesassociated functions of a prosthetic leg. Two users, one transtibial and one transfemoral,underwent surgery in order to get electrodes implantable into their residual leg muscles.They are among the rst users in the world to get this kind of implanted sensors.A prototype of the BSMB was manufactured. The process took more time thanexpected, mainly due to the fact that it was decided to use a ball grid array (BGA)microprocessor in order to save space. That meant more complicated routing and higherstandards for the manufacturing of the board. The results of the event detection indicatethat the data from the implanted electrodes can be used in order to get sucient controlover prosthetic legs. These are positive ndings for users of prosthetic legs and shouldincrease their security and quality of life.It is important to keep in mind when the results of this report are evaluated that allthe testing carried out were only done on one user each.

Automatically detecting events for people with diabetes mellitus using continuousglucose monitors is an important step in allowing insulin pumps to automaticallycorrect the blood glucose levels and for a more hands-off approach to thedisease. The automatic detection of events could also aid physicians whenassisting their patients when referring to their continuous glucose monitordata. A range of different deep learning algorithms has been applied forpredictions of different events for continuous glucose monitor data, such asthe onset for hyperglycemia, hypoglycemia or mealtime events. This thesisfocused on constructing sequences labelled from an unbalanced and assumedmisslabelled dataset to classify them as such using four different deep learningnetworks using convoluted neural networks and recurrent neural networks.Manual correction of the dataset allowed for only clear events starting witha high positive gradient to be labelled as positive. The classification wasperformed on exact timepoints and in time windows to allow the classificationto to be done around the beginning of an event instead of the exact timepoint.The results from using the unbalanced and assumed misslabelled datasetshowed the networks performing similarly, with high Recall and Precisionbelow 0.5, thus not found to be of use in a for automatic event detection.Further testing by using another dataset or further configurations is neededto clarify the capabilities of automatically detecting events. DDAnalytics willnot use any of the developed networks in any of their products. / Automatisk detection av event för personer med diabetes från deras kontinuerligaglukosmätare är ett viktigt steg för att låta insulinpumpar automatiskt korrigeraglukosnivåer och möjliggöra en mindre självreglering av personens diabetes.Denna automatiska detektion skulle även kunna hjälpa läkare vid samtalmed patienter och deras data från kontinuerliga glukosmätarna. En mängd avolika djupinlärningsalgoritmer har använts för förutsägelser av olika event förkontinuerlig glukosmätardata, som början av hyperglykemier, hypoglykemiereller måltider. Detta examensarbete fokuserar på skapandet av sekvenserfrån ett obalanserat och antaget inte helt korrekt markerade event i dataset,för att kunna klassificera dessa event med fyra olika djupinlärningsnätverk.Dessa nätverk bygger på konvolution och rekursiva neurala nätverk. Manuellkorrektion av datasetet möjliggjorde så att endast tydliga event som börjar meden kraftig positiv ökning av gradienten var markerade som positiva event.Klassificeringen genomfördes på både exakta tidssteg och i tidsfönster såatt början av event kunde detekteras snarare än bara det exakta tidssteget.Resultaten genom användandet av detta tidigare nämnda dataset visade liknanderesultat för samtliga nätverk, med hög Återkallelse och Precision under 0.5.Dessa resultat ledde till att nätverken inte kan antas kunna utföra automatiskevent detektion, och skulle behöva ytterligare testning på ett annat dataset medmer korrekta markerade event eller ytterligare konfigureringar på nätverken föratt verifiera dessas möjligheter att automatiskt klassificera event i kontinuerligglukosdata. DDanalytics kommer inte använda något av dessa framtagnanätverk i några av deras produkter.

The world that we live in is a complex network of agents and their interactions which are termed as events. An instance of an event is composed of directly measurable low-level actions (which I term sub-events) having a temporal order. Also, the agents can act independently (e.g. voting) as well as collectively (e.g. scoring a touch-down in a football game) to perform an event. With the dawn of the new millennium, the low-level vision tasks such as segmentation, object classification, and tracking have become fairly robust. But a representational gap still exists between low-level measurements and high-level understanding of video sequences. This dissertation is an effort to bridge that gap where I propose novel learning, detection, representation, indexing and retrieval approaches for multi-agent events in videos. In order to achieve the goal of high-level understanding of videos, firstly, I apply statistical learning techniques to model the multiple agent events. For that purpose, I use the training videos to model the events by estimating the conditional dependencies between sub-events. Thus, given a video sequence, I track the people (heads and hand regions) and objects using a Meanshift tracker. An underlying rule-based system detects the sub-events using the tracked trajectories of the people and objects, based on their relative motion. Next, an event model is constructed by estimating the sub-event dependencies, that is, how frequently sub-event B occurs given that sub-event A has occurred. The advantages of such an event model are two-fold. First, I do not require prior knowledge of the number of agents involved in an event. Second, no assumptions are made about the length of an event. Secondly, after learning the event models, I detect events in a novel video by using graph clustering techniques. To that end, I construct a graph of temporally ordered sub-events occurring in the novel video. Next, using the learnt event model, I estimate a weight matrix of conditional dependencies between sub-events in the novel video. Further application of Normalized Cut (graph clustering technique) on the estimated weight matrix facilitate in detecting events in the novel video. The principal assumption made in this work is that the events are composed of highly correlated chains of sub-events that have high conditional dependency (association) within the cluster and relatively low conditional dependency (disassociation) between clusters. Thirdly, in order to represent the detected events, I propose an extension of CASE representation of natural languages. I extend CASE to allow the representation of temporal structure between sub-events. Also, in order to capture both multi-agent and multi-threaded events, I introduce a hierarchical CASE representation of events in terms of sub-events and case-lists. The essence of the proposition is that, based on the temporal relationships of the agent motions and a description of its state, it is possible to build a formal description of an event. Furthermore, I recognize the importance of representing the variations in the temporal order of sub-events, that may occur in an event, and encode the temporal probabilities directly into my event representation. The proposed extended representation with probabilistic temporal encoding is termed P-CASE that allows a plausible means of interface between users and the computer. Using the P-CASE representation I automatically encode the event ontology from training videos. This offers a significant advantage, since the domain experts do not have to go through the tedious task of determining the structure of events by browsing all the videos. Finally, I utilize the event representation for indexing and retrieval of events. Given the different instances of a particular event, I index the events using the P-CASE representation. Next, given a query in the P-CASE representation, event retrieval is performed using a two-level search. At the first level, a maximum likelihood estimate of the query event with the different indexed event models is computed. This provides the maximum matching event. At the second level, a matching score is obtained for all the event instances belonging to the maximum matched event model, using a weighted Jaccard similarity measure. Extensive experimentation was conducted for the detection, representation, indexing and retrieval of multiple agent events in videos of the meeting, surveillance, and railroad monitoring domains. To that end, the Semoran system was developed that takes in user inputs in any of the three forms for event retrieval: using predefined queries in P-CASE representation, using custom queries in P-CASE representation, or query by example video. The system then searches the entire database and returns the matched videos to the user. I used seven standard video datasets from the computer vision community as well as my own videos for testing the robustness of the proposed methods.

Event extraction is a type of information extraction(IE) that works on extracting the specific knowledge of certain incidents from texts. Nowadays the amount of available information (such as news, blogs, and social media) grows in exponential order. Therefore, it becomes imperative to develop algorithms that automatically extract the machine-readable information from large volumes of text data. In this dissertation, we focus on three problems in obtaining event-related information from news articles. (1) The first effort is to comprehensively analyze the performance and challenges in current large-scale event encoding systems. (2) The second problem involves event detection and critical information extractions from news articles. (3) Third, the efforts concentrate on event-encoding which aims to extract event extent and arguments from texts. We start by investigating the two large-scale event extraction systems (ICEWS and GDELT) in the political science domain. We design a set of experiments to evaluate the quality of the extracted events from the two target systems, in terms of reliability and correctness. The results show that there exist significant discrepancies between the outputs of automated systems and hand-coded system and the accuracy of both systems are far away from satisfying. These findings provide preliminary background and set the foundation for using advanced machine learning algorithms for event related information extraction. Inspired by the successful application of deep learning in Natural Language Processing (NLP), we propose a Multi-Instance Convolutional Neural Network (MI-CNN) model for event detection and critical sentences extraction without sentence level labels. To evaluate the model, we run a set of experiments on a real-world protest event dataset. The result shows that our model could be able to outperform the strong baseline models and extract the meaningful key sentences without domain knowledge and manually designed features. We also extend the MI-CNN model and propose an MIMTRNN model for event extraction with distant supervision to overcome the problem of lacking fine level labels and small size training data. The proposed MIMTRNN model systematically integrates the RNN, Multi-Instance Learning, and Multi-Task Learning into a unified framework. The RNN module aims to encode into the representation of entity mentions the sequential information as well as the dependencies between event arguments, which are very useful in the event extraction task. The Multi-Instance Learning paradigm makes the system does not require the precise labels in entity mention level and make it perfect to work together with distant supervision for event extraction. And the Multi-Task Learning module in our approach is designed to alleviate the potential overfitting problem caused by the relatively small size of training data. The results of the experiments on two real-world datasets(Cyber-Attack and Civil Unrest) show that our model could be able to benefit from the advantage of each component and outperform other baseline methods significantly. / Ph. D. / Nowadays the amount of available information (such as news, blogs, and social media) grows in exponential order. The demand of making use of the massive on-line information during decision making process becomes increasing intensive. Therefore, it is imperative to develop algorithms that automatically extract the formatted information from large volumes of the unstructured text data. In this dissertation, we focus on three problems in obtaining event-related information from news articles. (1) The first effort is to comprehensively analyze the performance and challenges in current large-scale event encoding systems. (2) The second problem involves detecting the event and extracting key information about the event in the article. (3) Third, the efforts concentrate on extracting the arguments of the event from the text. We found that there exist significant discrepancies between the outputs of automated systems and hand-coded system and the accuracy of current event extraction systems are far away from satisfying. These findings provide preliminary background and set the foundation for using advanced machine learning algorithms for event related information extraction. Our experiments on two real-world event extraction tasks (Cyber-Attack and Civil Unrest) show the effectiveness of our deep learning approaches for detecting and extracting the event information from unstructured text data.