In this thesis, contributions are presented in the areas of shape parsing for view-based object recognition and spatio-temporal feature description for action recognition. A probabilistic model for parsing shapes into several distinguishable parts for accurate shape recognition is presented. This approach is based on robust geometric features that permit high recognition accuracy. As the second contribution in this thesis, a binary spatio-temporal feature descriptor is presented. Recent work shows that binary spatial feature descriptors are effective for increasing the efficiency of object recognition, while retaining comparable performance to state of the art descriptors. An extension of these approaches to action recognition is presented, facilitating huge gains in efficiency due to the computational advantage of computing a bag-of-words representation with the Hamming distance. A scene's motion and appearance is encoded with a short binary string. Exploiting the binary makeup of this descriptor greatly increases the efficiency while retaining competitive recognition performance.

Near-infrared spectroscopy (NIRS) in an imaging technique that has gained much attention in brain-computer interfaces (BCIs). Previous NIRS-BCI studies have primarily employed temporal features, derived from the time course of hemodynamic activity, despite potential value contained in the spatial attributes of a response. In an initial offline study, we investigated the value of using joint spatial-temporal pattern classification with dynamic NIR topograms to differentiate intentional cortical activation from rest. With the inclusion of spatiotemporal features, we demonstrated a significant increase in achievable classification accuracies from those obtained using temporal features alone (p < 10-4). In a second study, we evaluated the feasibility of implementing joint spatial-temporal pattern classification in an online system. We developed an online system-paced NIRS-BCI, and were able to differentiate two cortical states with high accuracy (77.4±10.5%). Collectively, these findings demonstrate the value of including spatiotemporal features in the classification of functional NIRS data for BCI applications.

This study aims to investigate the temporal-spatial distribution and reproduction of the Callionymidae, a dominant bottom-dwelling family at southwestern Taiwan. Samples were collected once every 1~2 month from January 2001 to January 2005 at seven stations, including Jiading, Zuoying, Jhongjhou, Linyuan, Dapeng Bay, Linbian and Fangliao, along the southwestern coast of Taiwan. A total of 5,846 samples was obtained, including 3 genera and 15 species. More species were found in this study than previous ones. The distribution of callionymids showed the significant variations in season, site and year. The highest abundance months occurred during March to October annually, with the most abundant at Jiading, then decreased in numbers southwards. Callionymus planus (52%)¡BCallionymus curvicornis (28%)¡BCallionymus virgis (7%) and Callionymus filamentosus (6%) were the top four dominant species. The four dominant species appeared abundantly around their spawning season. C. planus, C. curvicornis and C. filamentous were serial spawners, with peaking period at March-May, November-March and February-April, respectively. Both C. planus and C. curvicornis were most abundant at Jiading, and decreased southwardly, whereas the C. virgis and C. filamentosus were most abundant at Fangliao and Zuoying, respectively. All dominant species were revealed resource partitioning in relation to their reproductive activities. Callionymus planus that grew fast and recovered quickly among the callionymids, predominated in this area. Both C. planus and C. curvicornis showed significant reproductive isolation at the same area, both with a southward decrease in number. Calllionymus virgis separated from others and lived in the southmost site. Furthermore, C. filamentosus had a spawning period between C. planus and C. curvicornis.

Traditional fMRI utilizes blood oxygenation level dependent (BOLD) contrast to map brain activity. BOLD signal is sensitive to the hemodynamic changes associated with brain activity, and gives an indirect measure of brain activity. Low frequency fluctuations (LFFs) have been observed in the BOLD signal even in the absence of any anesthetic agent, and the correlations between the fluctuations from different brain regions has been used to map functional connectivity in the brain. Most studies involving spontaneous fluctuations in the BOLD signal extract connectivity patterns that show relationships between brain areas that are maintained over the length of the scanning session. The research presented in this document investigates the spatiotemporal dynamics of the BOLD fluctuations to identify common spatiotemporal patterns within a scan. First, the presence of a visually detectable spatiotemporal propagation pattern is demonstrated by utilizing single-slice data with high spatial and temporal resolution. The pattern consists of lateral-medial propagation of BOLD signal, demonstrating the presence of time-varying features in spontaneous BOLD fluctuations. Further, a novel pattern finding algorithm is developed for detecting repeated spatiotemporal patterns in BOLD fMRI data. The algorithm is applied to high temporal resolution T2*-weighted multislice images obtained from rats and humans in the absence of any task or stimulation. In rats, the primary pattern consists of waves of high signal intensity, propagating in a lateral-medial direction across the cortex, replicating the results obtained using visual observation. In humans, the most common spatiotemporal pattern consisted of an alteration between activation of areas comprising the "default-mode" (e.g., posterior cingulate and anterior medial prefrontal cortices) and the "task-positive" (e.g., superior parietal and premotor cortices) networks. Signal propagation from focal starting points is also observed. The pattern finding algorithm is shown to be reasonably insensitive to the variation in user-defined parameters, and the results are consistent within and between subjects. This novel approach for probing the spontaneous network activity of the brain has implications for the interpretation of conventional functional connectivity studies, and may increase the amount of information that can be obtained from neuroimaging data.

Central New York is located in one of the snowiest regions in the United States, with the city of Syracuse, New York the snowiest metropolis in the nation. Snowfall in the region generally begins in mid-November and lasts until late-March. Snow accumulation occurs from a multitude of conditions: frontal systems, mid-latitude cyclones, Nor'easters, and most notably lake-effect storms. Lake effect snowfall (LES) is a difficult parameter to forecast due to the isolated and highly variable nature of the storm. Consequently, studies have attempted to determine changes in snowfall for lake-effect dominated regions. Annual snowfall patterns are of particular concern as seasonal snowfall totals are vital for water resources, winter businesses, agriculture, government and state agencies, and much more. Through the use of snowfall, temperature, precipitation, and location data from the National Weather Service's Cooperative Observer Program (COOP), spatial and temporal changes in snowfall for Central New York were determined. In order to determine climatic changes in snowfall, statistical analyses were performed (i.e. least squares estimation, correlations, principal component analyses, etc.) and spatial maps analyzed. Once snowfall trends were determined, factors influencing the trends were examined. Long-term snowfall trends for CNY were positive for original stations (~0.46 +/- 0.20 in. yr-1) and homogenously filtered stations (0.23 +/- 0.20 in. yr-1). However, snowfall trends for shorter time-increments within the long-term period were not consistent, as positive, negative, and neutral trends were calculated. Regional differences in snowfall trends were observed for CNY as typical lake-effect areas (northern counties, the Tug Hill Plateau and the Southern Hills) experienced larger snowfall trends than areas less dominated by LES. Typical lake-effect months (December - February) experienced the greatest snowfall trend in CNY compared to other winter months. The influence of teleconnections on seasonal snowfall in CNY was not pronounced; however, there was a slight significant (5%) correlation (< 0.35) with the Atlantic Multidecadal Oscillation. It was not clear if changes in air temperature or changes in precipitation were the cause of variations in snowfall trends. It was also inconclusive if the elevation or distance from Lake Ontario resulted in increased snowfall trends. Results from this study will aid in seasonal snowfall forecasts in CNY, which can be used to predict future snowfall. Even though the study area is regionally specific, the methods may be applied to other lake effect dominated areas to determine temporal and spatial variations in snowfall. This study will enhance climatologists and operational forecasters' awareness and understanding of snowfall, especially lake effect snowfall in CNY.

Physical and chemical heterogeneities cause the porous media transport parameters to vary with scale, and between these two types of heterogeneities geological heterogeneity is considered to be the most important source of scale-dependence of transport parameters. Subsurface processes associated with chemical alterations result in changing reservoir properties with interlinked spatial and temporal scale, and there is uncertainty in the evolution of those properties and the chemical processes. This dissertation provides a framework and procedures to quantify the spatiotemporal scaling characteristics of reservoir attributes and transport processes in heterogeneous media accounting for chemical alterations in the reservoir. Conventional flow scaling groups were used to assess their applicability in scaling of recovery and Mixing Zone Length (MZL) in presence of chemical reactivity and permeability heterogeneity through numerical simulations of CO₂ injection. It was found out that these scaling groups are not adequate enough to capture the scaling of recovery and transport parameters in the combined presence of chemical reactivity and physical heterogeneity. In this illustrative example, MZL was investigated as a function of spatial scale, temporal scale, multi-scale heterogeneity, and chemical reactivity; key conclusions are that 1) the scaling characteristics of MZL distinctly differ for low permeability and high permeability media, 2) heterogeneous media with spatial arrangements of both high and low permeability regions exhibit scaling characteristics of both high and low permeability media, 3) reactions affect scaling characteristics of MZL in heterogeneous media, 4) a simple rescaling can combine various MZL curves by merging them into a single MZL curve irrespective of the correlation length of heterogeneity, and 5) estimates of MZL (and consequently predictions of oil recovery) will fluctuate corresponding to displacements in a permeable medium whose lateral length is smaller than the correlation length of geological formation. We illustrate and extend the procedure of estimating Representative Elementary Volume (REV) to include temporal scale by coupling it with spatial scale. The current practice is to perform spatial averaging of attributes and account for residual variability by calibration and history matching. This results in poor predictions of future reservoir performance. The proposed semi-analytical technique to scale-up in both space and time provides guidance for selection of spatial and temporal discretizations that takes into account the uncertainties due to sub-processes. Finally, a probabilistic particle tracking (PT) approach is proposed to scale-up flow and transport of diffusion-reaction (DR) processes while addressing multi-scale and multi-physics nature of DR mechanisms and also maintaining consistent reservoir heterogeneity at different levels of scales. This multi-scale modeling uses a hierarchical approach which is based on passing the macroscopic subsurface heterogeneity down to the finer scales and then returning more accurate reactive flow response. This PT method can quantify the impact of reservoir heterogeneity and its uncertainties on statistical properties such as reaction surface area and MZL, at various scales. / text

The mammalian brain is a complex computing system that contains billions of neurons and trillions of connections. Is there a general principle that governs the processing in such large neural populations? This dissertation attempts to address this question using computational modeling and quantitative analysis of direct physiological measurements of large neural populations in the monkey primary visual cortex (V1). First, the complete spatiotemporal dynamics of V1 responses over the entire region that is activated by small stationary stimuli are characterized quantitatively. The dynamics of the responses are found to be systematic but complex. Importantly, they are inconsistent with many popular computational models of neural processing. Second, a simple population gain control (PGC) model that can account for these complex response properties is proposed for the small stationary stimuli. The PGC model is then used to predict the responses to stimuli composed of two elements and stimuli that move at a constant speed. The predictions of the model are consistent with the measured responses in V1 for both stimuli. PGC is the first model that can account for the complete spatiotemporal dynamics of V1 population responses for different types of stimuli, suggesting that gain control is a general mechanism of neural processing. / text

Η εξόρυξη χώρο-χρονικών δεδομένων αποτελεί πλέον μία από τις σημαντικότερες κατευθύνσεις του κλάδου της εξόρυξης γνώσης. Κάποια από τα βασικά προβλήματα που καλείται να αντιμετωπίσει είναι η ανακάλυψη περιοχών που εμφανίζουν ομοιότητες στην χρονική τους εξέλιξη, η αναγνώριση προτύπων που εμφανίζονται τόσο στην χωρική όσο και στη χρονική πληροφορία, η πρόβλεψη μελλοντικών τιμών και η αποθήκευση σε εξειδικευμένες βάσεις δεδομένων με σκοπό την αποδοτική απάντηση χωροχρονικών ερωτημάτων. Οι μέθοδοι που προσεγγίζουν τα παραπάνω προβλήματα καθώς και οι βασικές εργασίες της εξόρυξης γνώσης, όπως η κατηγοριοποίηση και η ομαδοποίηση, εμφανίζονται στον πυρήνα της πλειονότητας των εργαλείων ανάλυσης και επεξεργασίας χώρο-χρονικών δεδομένων. Βασικός στόχος της παρούσας εργασίας είναι η εφαρμογή μεθόδων εξόρυξης χώρο-χρονικών δεδομένων στο Ηλεκτροεγκεφαλογράφημα (ΗΕΓ), το οποίο αποτελεί μία από τις πιο διαδεδομένες τεχνικές ανάλυσης της εγκεφαλικής λειτουργίας. Τα δεδομένα που προκύπτουν από το ΗΕΓ περιέχουν τόσο χωρική όσο και χρονική πληροφορία καθώς αποτελούνται από ηλεκτρικά σήματα που προέρχονται από ηλεκτρόδια τοποθετημένα σε συγκεκριμένες θέσεις στο κρανίο. Τα βασικά προβλήματα που μελετήθηκαν στην επεξεργασία του ΗΕΓ είναι η μοντελοποίηση και η συσταδοποίηση χώρο-χρονικών δεδομένων, τα οποία οδήγησαν στην ανάπτυξη των αντίστοιχων μεθόδων. Στα πλαίσια της παρούσας εργασίας μελετήθηκε επίσης το πρόβλημα της διαχείρισης των δεδομένων ΗΕΓ και τη ανάλυσης ροών δεδομένων σε πραγματικό χρόνο. Η ενασχόληση με τα συγκεκριμένα προβλήματα οδήγησε α) στη δημιουργία καινοτόμων μεθόδων μοντελοποίησης και συσταδοποίησης χωρο-χρονικών δεδομένων, β) στον σχεδιασμό μιας βάσης δεδομένων, γ) στην μελέτη της βιβλιογραφίας στο θέμα της εξόρυξης και της διαχείρισης ροών δεδομένων και δ) στην δημιουργία μιας εφαρμογής για την ανάλυση δεδομένων σε πραγματικό χρόνο πάνω σε ένα σύστημα διαχείρισης ροών δεδομένων. Η παρούσα εργασία περιλαμβάνει ένα ένα σύνολο μεθόδων και εργαλείων ανάλυσης και διαχείρισης δεδομένων που εξετάστηκαν και χρησιμοποιήθηκαν προκειμένου να μελετηθεί η καταλληλότητά της εφαρμογής τους στις καταγραφές ΗΕΓ. Με τον τρόπο αυτό επιτυγχάνεται ο πρωταρχικός στόχος της εργασίας: η προώθηση υπαρχόντων και η δημιουργία καινοτόμων μεθόδων ανάλυσης από τον κλάδο της εξόρυξης γνώσης στα δεδομένα του ηλεκτροεγκεφαλογραφήματος. / Mining spatiotemporal data is one of the most significant topics in the field of data mining and knowledge discovery. Detecting locations that exhibit similarities in their temporal evolution, recognizing patterns that appear in both spatial and temporal information and storing spatiotemporal data in specialized databases are some of the fundamental problems tackled by researchers in this specific area. Methods and algorithms that address such problems along with the common data mining tasks (e.g. classification and clustering) are critical in the development of applications for analyzing spatiotemporal data, fact that highlights the necessity of continuous advancements of these algorithms in terms of usability, accuracy and performance. The most significant objective of the work performed during this thesis is the application of spatiotemporal data mining methods on the analysis of EEG, in order to exploit the both the spatial and the temporal nature of these data (i.e. electrodes placed on specific locations on the scalp that continuously record the electrical activity of the brain). Towards this direction the problems of modeling and clustering spatiotemporal data were extensively studied and the major outcome was the development of two corresponding methods. Furthermore, during this work the problem of managing EEG data was investigated both in the offline and the online scenario and within the latter, the state of the art in mining data streams was studied. The outcomes of this thesis related to the aforementioned problems include a) the development of a graph-based method for modeling spatiotemporal data, b) a method for clustering spatiotemporal data based on this model, c) the design of a database schema for storing eeg recording data and meta-data and d) the development of an application for online spindle detection over a data stream management system. Finally, this work aims towards the development of new and the adaptation of existing data mining methods in the context of spatiotemporal EEG analysis.

In this thesis, contributions are presented in the areas of shape parsing for view-based object recognition and spatio-temporal feature description for action recognition. A probabilistic model for parsing shapes into several distinguishable parts for accurate shape recognition is presented. This approach is based on robust geometric features that permit high recognition accuracy. As the second contribution in this thesis, a binary spatio-temporal feature descriptor is presented. Recent work shows that binary spatial feature descriptors are effective for increasing the efficiency of object recognition, while retaining comparable performance to state of the art descriptors. An extension of these approaches to action recognition is presented, facilitating huge gains in efficiency due to the computational advantage of computing a bag-of-words representation with the Hamming distance. A scene's motion and appearance is encoded with a short binary string. Exploiting the binary makeup of this descriptor greatly increases the efficiency while retaining competitive recognition performance.

PhotoScope, a novel visualization, visualizes the spatiotemporal coverage of photos in a construction photo collection. It extends the standard photo browsing paradigm in two main ways: visualizing spatial coverage of photos on floor plans, and indexing photos by a combination of spatial coverage, time, and content specifications. This approach enables users to browse and search space- and time-indexed photos more effectively. We designed PhotoScope specifically to address challenges in the construction management industry, where large photo collections are amassed to document project progress. These ideas may also apply to any photo collection that is spatially constrained and must be searched using spatial, temporal, and content criteria. Design choices made when developing PhotoScope are also described. Civil, mechanical and electrical engineers, and professionals from construction management validated the visualization mechanisms and functionalities of PhotoScope in a usability study. Empirical findings on the cognitive behaviors of participants are also discussed in this thesis.