Temporal Stream Research Articles

Estimating Temporal Dynamics of Human Emotions

Sentiment analysis predicts a one-dimensional quantity describing the positive or negative emotion of an author. Mood analysis extends the one-dimensional sentiment response to a multi-dimensional quantity, describing a diverse set of human emotions. In this paper, we extend sentiment and mood analysis temporally and model emotions as a function of time based on temporal streams of blog posts authored by a specific author. The model is useful for constructing predictive models and discovering scientific models of human emotions.

Inferring action structure and causal relationships in continuous sequences of human action

In the real world, causal variables do not come pre-identified or occur in isolation, but instead are embedded within a continuous temporal stream of events. A challenge faced by both human learners and machine learning algorithms is identifying subsequences that correspond to the appropriate variables for causal inference. A specific instance of this problem is action segmentation: dividing a sequence of observed behavior into meaningful actions, and determining which of those actions lead to effects in the world. Here we present a Bayesian analysis of how statistical and causal cues to segmentation should optimally be combined, as well as four experiments investigating human action segmentation and causal inference. We find that both people and our model are sensitive to statistical regularities and causal structure in continuous action, and are able to combine these sources of information in order to correctly infer both causal relationships and segmentation boundaries.

Spacetime Stereo and 3D Flow via Binocular Spatiotemporal Orientation Analysis.

This paper presents a novel approach to recovering estimates of 3D structure and motion of a dynamic scene from a sequence of binocular stereo images. The approach is based on matching spatiotemporal orientation distributions between left and right temporal image streams, which encapsulates both local spatial and temporal structure for disparity estimation. By capturing spatial and temporal structure in this unified fashion, both sources of information combine to yield disparity estimates that are naturally temporal coherent, while helping to resolve matches that might be ambiguous when either source is considered alone. Further, by allowing subsets of the orientation measurements to support different disparity estimates, an approach to recovering multilayer disparity from spacetime stereo is realized. Similarly, the matched distributions allow for direct recovery of dense, robust estimates of 3D scene flow. The approach has been implemented with real-time performance on commodity GPUs using OpenCL. Empirical evaluation shows that the proposed approach yields qualitatively and quantitatively superior estimates in comparison to various alternative approaches, including the ability to provide accurate multilayer estimates in the presence of (semi)transparent and specular surfaces.

Efecto de los factores ambientales sobre la diversidad de insectos hemimetábolos y coleópteros acuáticos en la cuenca del Río Xichú, Guanajuato, México

Lotic systems, in particular temporal streams, represent an important proportion of continental waters, but are poorly studied. This is particularly clear in central Mexico, where temporal streams are abundant and remain poor studied despite their great potential for high diversity and the important energy within these ecosystems. The aim of this study was to analyze the environmental variables and scales at which these variables have effects on aquatic insect assemblages in a semiarid fluvial system in Central Mexico. A total of 89 genera were registered, the order Coleoptera showing the highest richness. Peak values for alpha diversity were found on permanent sites in the dry season (1D = 10.63), the more stable environment in permanent streams during the dry season might accounted for this pattern. Assemblages at our study sites showed low similarity (Ij < 0.45) indicating high genera replacement among sites, possibly related to differences in hydroperiod. Partition analysis indicates that variation in genera composition among rivers contributed a large proportion of the diversity (37%). Results from the canonical correspondence analysis point to mesohabitat, velocity, ecological quality, and hydrological order, as the most important variables explaining diversity patterns. In addition, hydrological condition and intermittency were important factors related to diversity (-21-22%). The results suggest that permanent sites function has species sources for intermittent sites, highlighting the need for conservation planning at the landscape level in the Xichu's basin.

Continuous judgment by category-ratio scaling of aircraft noise

A continuous judgment by category-ratio (CJCR) scaling method was used to evaluate the instantaneous annoyance of original and modified aircraft sounds. The result from the proposed method resulted in a temporal stream of annoyance levels for the whole flyover sequence that could be further analyzed. The test subjects were continuously rating their instantaneous annoyance on a Borg CR 100 scale® during the playback of 10 flyover sequences. Using a category-ratio (CR) scale instead of a category (C) scale, mathematical operations such as calculations of average were enabled but gave also advantages in terms of higher resolution in the responses. The results showed differences in perception in the time segment where the sound had been modified. The temporal stream of annoyance was also converted into overall judgments of the sounds, these estimations showed consistency with previous results obtained using the semantic differential and paired comparison method.

Spatiotemporal flow of information in the early visual pathway

The spatial components of a visual scene are processed neurally in a sequence of coarse features followed by fine features. This coarse-to-fine temporal stream was initially considered to be a cortical function, but has recently been demonstrated in the dorsal lateral geniculate nucleus. The goal of this study was to test the hypothesis that coarse-to-fine processing is present at earlier stages of visual processing in the retinal ganglion cells that supply lateral geniculate nucleus (LGN) neurons. To compare coarse-to-fine processing in the cat's visual system, we measured the visual responses of connected neuronal pairs from the retina and LGN, and separate populations of cells from each region. We found that coarse-to-fine processing was clearly present at the ganglion cell layer of the retina. Interestingly, peak and high-spatial-frequency cutoff responses were higher in the LGN than in the retina, indicating that there was a progressive cascade of coarse-to-fine information from the retina to the LGN to the visual cortex. The analysis of early visual pathway receptive field characteristics showed that the physiological response interplay between the center and surround regions was consistent with coarse-to-fine features and may provide a primary role in the underlying mechanism. Taken together, the results from this study provided a framework for understanding the emergence and refinement of coarse-to-fine processing in the visual system.

Self-Induced Attentional Blink

Satisfaction of search (which we refer to as subsequent search misses)—a decrease in accuracy at detecting a second target after a first target has been found in a visual search—underlies real-world search errors (e.g., tumors may be missed in an X-ray if another tumor already has been found), but little is known about this phenomenon’s cognitive underpinnings. In the present study, we examined subsequent search misses in terms of another, more extensively studied phenomenon: the attentional blink, a decrease in accuracy when a second target appears 200 to 500 ms after a first target is detected in a temporal stream. Participants searched for T-shaped targets among L-shaped distractors in a spatial visual search, and despite large methodological differences between self-paced spatial visual searches and attentional blink tasks, an attentional-blink-like effect accounted for subsequent-search-miss errors. This finding provides evidence that accuracy is negatively affected shortly after a first target is fixated in a self-paced, self-guided visual search.

Music Cognition as Mental Time Travel

As we experience a temporal flux of events our expectations of future events change. Such expectations seem to be central to our perception of affect in music, but we have little understanding of how expectations change as recent information is integrated. When music establishes a pitch centre (tonality), we rapidly learn to anticipate its continuation. What happens when anticipations are challenged by new events? Here we show that providing a melodic challenge to an established tonality leads to progressive changes in the impact of the features of the stimulus on listeners' expectations. The results demonstrate that retrospective analysis of recent events can establish new patterns of expectation that converge towards probabilistic interpretations of the temporal stream. These studies point to wider applications of understanding the impact of information flow on future prediction and its behavioural utility.

Neural architectures for learning to answer questions

Abstract Two related and integral parts of linguistic competence are the ability to comprehend incoming utterances and the ability to respond to them appropriately. In this context, we present two biologically inspired recurrent neural network models, based on the long short-term memory (LSTM) architecture, each of which develops a functional grasp of a small language by participating in a question–answering game. Both models receive questions in English, presented as temporal streams of speech sound patterns. As a further input, the models receive a set of symbolic facts about a simulated visual environment. The models learn to correctly answer input questions by observing question–answer pairs produced by other participants. The first of our two models produces its answers as symbolic facts, demonstrating an ability to ground language. The second model learns by observation to produce its answers as full English sentences. This latter task in particular is closely analogous to that faced by human language learners, involving segmentation of morphemes, words, and phrases from observable auditory input, mapping of speech signals onto intended environmental referents, comprehension of questions, and content-addressed search capabilities for discovering the answers to these questions. Analysis of the models shows that their performance depends upon highly systematic learned representations that combine the best properties of distributed and symbolic representations.

Lorine Niedecker, Henri Bergson and the Poetics of Temporal Flow

While Niedecker's “life by water” is the subject of many of her poems, reflecting the experience of living on the flood-prone island of Black Hawk, Wisconsin, the water imagery that saturates many of her poems also becomes a way of reflecting upon time as it is experienced, of consciousness as process, of subjectivity as something pulled along by a temporal stream. The poetics of temporal flow suggests Niedecker's interest in exploring forms of experience that fall outside conscious attention and that point to her affinity with the modernist philosopher of time, Henri Bergson, whoseCreative Evolutionshe read in 1955. This paper examines some of the short, concise poems Niedecker wrote in the 1950s, together with the later, looser-limbed poems such as “Paean to Place” where she turns away from the objectivist principles of com pression in order to produce fluid, malleable poems commensurate with the flow of experience.

Performance Modeling of Spatio-Temporal Algorithms Over GEDS Framework

The efficient processing of spatio-temporal data streams is an area of intense research. However, all methods rely on an unsuitable processor (Govindaraju, 2004), namely a CPU, to evaluate concurrent, continuous spatio-temporal queries over these data streams. This paper presents a performance model of the execution of spatio-temporal queries over the authors’ GEDS framework (Cazalas &amp; Guha, 2010). GEDS is a scalable, Graphics Processing Unit (GPU)-based framework, employing computation sharing and parallel processing paradigms to deliver scalability in the evaluation of continuous, spatio-temporal queries over spatio temporal data streams. Experimental evaluation shows the scalability and efficacy of GEDS in spatio-temporal data streaming environments and demonstrates that, despite the costs associated with memory transfers, the parallel processing power provided by GEDS clearly counters and outweighs any associated costs. To move beyond the analysis of specific algorithms over the GEDS framework, the authors developed an abstract performance model, detailing the relationship of the CPU and the GPU. From this model, they are able to extrapolate a list of attributes common to successful GPU-based applications, thereby providing insight into which algorithms and applications are best suited for the GPU and also providing an estimated theoretical speedup for said GPU-based applications.

An analytical solution for a radial collector well near a stream with a low-permeability streambed

Summary Radial collector wells are often constructed near a stream to obtain more water and produce smaller drawdown in comparison with traditional vertical wells. This paper aims at developing a mathematical model for describing head distributions induced by pumping in a radial collector well near a stream in an unconfined aquifer. A low permeable streambed between the stream and the aquifer is considered. The first-order free surface equation is used to describe the movement of water table. A point-sink solution of the model is developed first by a general transform combining Fourier sine and cosine transforms and then by Fourier transform and Laplace transform. The transient solution of head distributions for different well types such as the horizontal well or radial collector well can be obtained by integrating the point-sink solution along the well. Based on Darcy’s law and the developed solution, an equation for temporal stream depletion rate describing filtration from the stream is then obtained. The steady-state solutions for filtration can be obtained from the transient solution when neglecting the exponential term associated with time. It is found that steady-state filtration depends only on the ratio of streambed permeability over aquifer permeability. Steady-state filtration equals water extraction from a well when the ratio is larger than 10−2. The streambed is regarded as completely impermeable when the ratio is less than 10−7. Additionally, the lowest water table happens near the well before occurrence of filtration. The lowest water table moves landward and away from the well after filtration occurs.

Machine Learning for Traffic Prediction

Recently, GPS car navigation became popular enough for a significant number of drivers in any large city to be able to potentially send their current GPS positions and velocities to a central traffic-managing server in real time. This data could be used to reconstruct a map of current traffic in the whole city and make predictions for the future, which would be valuable for traffic control, congestion analysis and prevention. The difficulty lying in the traffic reconstruction problem is that the number of cars equipped with a GPS transmitter is a small in comparison with the total number of cars in the road network. Information obtained from GPS transmitters is therefore rare and incomplete, raising the problem of approximating travel velocities from sparse information. In this context, the paper proposes approaches to solving two problems: (1) of reconstruction of routes of the individual cars in a street graph from a temporal stream of its coordinates and (2) of approximating the average velocity on a given street from irregular time series of instantaneous velocities of cars passing through that street. Determining the car's position in the road network is quite expensive computationally when done in a naive way, so we propose a method to speed up the car's localization based on the R-tree data structure [10]. Additionally, a resilient propagation neural network [13] is applied to approximate the average velocity on any edge of a street graph. Our methods were applied to one of the problems announced in the competition organized in June 2010 by Tom-Tom - a company producing automotive navigation systems [15]. The results of experiments show the capacity of our approach to make useful predictions.

Distinct structural plasticity in the hippocampus and amygdala of the middle-aged common marmoset ( Callithrix jacchus)

Adult neurogenesis in the primate brain is generally accepted to occur primarily in two specific areas; the subgranular zone (SGZ) of the hippocampal dentate gyrus (DG) and the subventricular zone (SVZ) of the lateral ventricles. Hippocampal neurogenesis is well known to be downregulated by stress and aging in rodents, however there is less evidence documenting the sensitivity of neuroblasts generated in the SVZ. In primates, migrating cells generated in the SVZ travel via a unique temporal stream (TS) to the amygdala and entorhinal cortex. Using adult common marmoset monkeys ( Callithrix jacchus), we examined whether i) adult-generated cells in the marmoset amygdala differentiate into doublecortin-positive (DCX+) neuroblasts, and ii) whether lasting changes occur in DCX-expressing cells in the DG or amygdala when animals are exposed to 2 weeks of psychosocial stress. A surprisingly large population of DCX+ immature neurons was found in the amygdala of these 4-year-old monkeys with an average density of 163,000 DCX+ cells per mm 3. Co-labeling of these highly clustered cells with PSA-NCAM supports that a subpopulation of these cells are migratory and participate in chain-migration from the SVZ to the amygdala in middle-aged marmosets. Exposure to 2 weeks of isolation and social defeat stress failed to alter the numbers of BrdU+, or DCX+ cells in the hippocampus or amygdala when evaluated 2 weeks after psychosocial stress, indicating that the current stress paradigm has no long-term consequences on neurogenesis in this primate.

Application of SWAT model to small agricultural catchment in Poland

Poland is obliged, like the other EU countries, to implement the Water Framework Directive - WFD (2000/60/WE) by the end of 2015. The main objective of WFD is to provide normative quality of all water resources. To reach this goal reduction of water polluter emission to the environment is needed. Our project focuses on pollution from agricultural sources which share in global pollution is high and growing still. As a pilot area, where the WFD is going to be implemented, small agricultural Zglowiączka catchment was chosen.The state monitoring of surface water quality for the catchment is conducted in three points along the Zglowiaczka River. In each of these three points, nitrates concentration periodically significantly exceeds the allowable value of 50 mg NO₃·dm⁻³. The highest average monthly values of nitrates concentration in years 1990-2007 occur in February, March and April, which indicates on agriculture as a source of pollution. The Zglowiaczka catchment is an area where reduction of nitrogen run-off from agricultural lands to water resources is especially needed. The main topic of the research carried out in the Polish-Norwegian project is to propose different means for reduction of migration of nitrate to surface water based on modeling approach. In the paper a conception of creating buffer zones using SWAT model is presented. We considered fitting the buffer zone width, depending on the flow rate of water flowing from the fields to the stream. Using SWAT model interface a map of potential flow under the conditions of the intensive precipitation was generated. The next step was distribution over the whole Zglowiączka catchment, places with high of the temporal streams network. It was done using GRASS program. The map of stream density was done by assigning the raster number which is the sum of raster in the neighbourhood (radius of neighbourhood smaller or equal 25 raster). The choice of the most endangered subbasins was done on base of visual evaluation of the surface flow map.It is visible in the results that filter strips on endangered areas are far more effective and therefore more required. If the width of the vegetated buffer strips is not sufficient, it will not attain the desired effectiveness. Conversely, if the width is too great, it will cause agricultural land waste, preventing farmers' interest in cooperating with environmental preservation efforts. For the above reasons, it is important to set a reasonable width range. According to the results we are suggesting wider buffer zones in endangered subbasins and narrow in other subbasins.

Alerting enhances target identification but does not affect the magnitude of the attentional blink

Identification of the second of two targets is impaired when presented less than about 500 ms after the first. The magnitude of this attentional blink (AB) is known to be modulated by tonic factors (e.g., observer's state of relaxation). The present work examined the effects of a phasic change in observer's state brought about by an alerting stimulus (an aggregate of faint rings) presented in temporal proximity to either letter-target inserted in a temporal stream (RSVP) of digit distractors. In four experiments, identification accuracy of each target was substantially improved by presenting the alerting stimulus either in the target's frame or in the preceding RSVP frame. However, alerting did not modulate the magnitude of the AB. The appearance of an alerting effect on the AB in Experiment 1 was ascribed to a ceiling effect in Experiment 2. Experiment 3 ruled out endogenous temporal cueing effects; Experiment 4 examined the temporal gradient of alerting. Independence of the alerting and AB effects suggests that the alerting stimuli and the letter targets may be processed along distinct visual pathways.

Identifying predictive multi-dimensional time series motifs: an application to severe weather prediction

We introduce an efficient approach to mining multi-dimensional temporal streams of real-world data for ordered temporal motifs that can be used for prediction. Since many of the dimensions of the data are known or suspected to be irrelevant, our approach first identifies the salient dimensions of the data, then the key temporal motifs within each dimension, and finally the temporal ordering of the motifs necessary for prediction. For the prediction element, the data are assumed to be labeled. We tested the approach on two real-world data sets. To verify the generality of the approach, we validated the application on several subjects from the CMU Motion Capture database. Our main application uses several hundred numerically simulated supercell thunderstorms where the goal is to identify the most important features and feature interrelationships which herald the development of strong rotation in the lowest altitudes of a storm. We identified sets of precursors, in the form of meteorological quantities reaching extreme values in a particular temporal sequence, unique to storms producing strong low-altitude rotation. The eventual goal is to use this knowledge for future severe weather detection and prediction algorithms.

Temporal and Social Context Based Burst Detection from Folksonomies

Burst detection is an important topic in temporal stream analysis. Usually, only the textual features are used in burst detection. In the theme extraction from current prevailing social media content, it is necessary to consider not only textual features but also the pervasive collaborative context, e.g., resource lifetime and user activity. This paper explores novel approaches to combine multiple sources of such indication for better burst extraction. We systematically investigate the characters of collaborative context, i.e., metadata frequency, topic coverage and user attractiveness. First, a robust state based model is utilized to detect bursts from individual streams. We then propose a learning method to combine these burst pulses. Experiments on a large real dataset demonstrate the remarkable improvements over the traditional methods.

A dynamical model of perceptual grouping

In a visual scene, we perceive similar visual objects as being grouped together. We propose a high-level, mechanistic model of perceptual grouping where stimuli that are sufficiently similar are dynamically clustered forming a visual category. The model assumes that each stimulus can be characterized by a continuous feature, like an orientation, which is encoded in the neuronal activity of a recurrent network. The connectivity of the network is sufficiently structured to allow for the generation of feature-selective stable states, or bumps, while maintaining the stability of a non-specific, low-activity state associated with spontaneous activity. The emergence of bumps corresponds to the formation of stimulus categories and is triggered by external inputs, idealized as temporal sequences of feature values. We suppose that each feature value in the sequence is sub-threshold; it leads to a response that, compared to the stimulus presentation rate, decays slowly after the stimulus is removed. When the sequence of stimuli is presented, the network sums temporally the responses triggered separately by each item. Only if the input stream consists of an adequate number of similar feature values, will there be successive hits within a critical time window in one or more localized regions of feature space that will give rise to the emergence of bump activity states. The network encodes therefore the distribution of feature values of the temporal input stream and, by extension, the visual scene.

Effects of Phragmites australis growth on nitrogen retention in a temporal stream

En los ultimos anos, en el SE iberico con el incremento de la superficie de regadio, se ha producido un crecimiento masivo de las poblaciones de Phragmites australis que terminan invadiendo por completo las ramblas y en general los cauces de aguas superficiales. Esta situacion lleva consigo la transformacion fisica de los cauces, modificando muchas caracteristicas implicadas en los procesos bioticos y abioticos de retencion de N. En este estudio testamos la hipotesis de que la ocupacion de los cauces por Phragmites australis afecta negativamente a la retencion de N. Asi comparamos las tasas de retencion (%R) de N - NO-3 y N - NH+4 en diferentes subtramos de una misma rambla: un subtramo no vegetado (238 m2) y dos subtramos vegetados que diferian en superficie (480 m2 y 910 m2). Los resultados demostraron que las eficacias de retencion (%R) para ambos solutos fueron superiores en el tramo no vegetado. Aunque no existen resultados concluyentes parece intuirse que estas diferencias fueron mas acusadas fuera del periodo vegetativo del helofito, mientras que en los meses de primavera y verano pudiera ocurrir un incremento de las tasas de retencion en los tramos vegetados. Asi mismo, la capacidad de los subtramos para la retencion de N-nitrato, mostro una clara dependencia de los aportes de nitrogeno, disminuyendo conforme la carga de nitrogeno aumenta. Sin embargo, el tramo no vegetado mostro una mayor capacidad de carga que el tramo vegetado de mayor superficie y este que el tramo vegetado de menor superficie. Este estudio pone de manifiesto que la ocupacion de los cauces por Phragmites australis, fenomeno generalizado en muchas partes del mundo, no solo puede suponer cambios en la estructura de la vegetacion y la fauna de las ramblas sino que tambien afectar a su funcionamiento y muy especialmente a un proceso clave implicado en la calidad de las aguas como es la eliminacion del nitrogeno.