Search

While there are many kinds of real-time systems in use, in production, or planned that have requirements for real-time database functionality, there is little common understanding of the underlying requirements that lead to such implementations. This situation extends both to the implementation community which must produce these systems as well as to the research community seeking to understand the fundamental principles that describe their behavior.

This paper presents a practical implementation of a person verification system based on face profiles in a cooperative environment. The image acquisition constraints are given and motivated. A set of global and local features are automatically extracted from the profile outline of the head. The candidate profile is then compared with profiles of the database to issue a distance from which to accept or reject the claimed individual. To increase robustness, the temporal consistency of the decisions over several shots is examined, what is possible thanks to the overall speed of the developed system.

We investigated the 30-min. temporal consistency and temporal stability of three focal scales for measuring stress: the Cognitive Anxiety Scale-State, the Focusing Inventory Scale-State, and the Somatic Symptom Scale-State. These scales are applicable for individually measured reduction in stress in research on outcome with older adolescents and adults. All possess excellent internal consistency. All manifested significant temporal consistency and good to excellent temporal stability over 30 min. in a nonclinical sample of 60 undergraduates. Small significant differences in means on the Cognitive Anxiety Scale-State and the Somatic Symptom Scale-State were noted. The phenomenological states associated with stress were stable and consistent over 30 min., the optimal time interval used in research on stress.

The recent advances in multimedia systems, together with the advent of high speed networks, paved the way to a new generation of applications. In particular, the authoring environments have found in multimedia the means of increasing the richness of information contained in electronic documents. With the evolution of new computer systems that can handle multimedia information, time-based data can be integrated in electronic documents taking into account their temporal dimension. In such documents, temporal dependencies between the different media objects define a temporal structure within the document. This structure is the basic support for the representation of dependencies between data such as audio, video and virtual images. Furthermore, it allows the scheduling of presentation actions during the document presentation. The presentation of multimedia documents is dynamic and the positioning of objects in time together with their duration have to be specified. To achieve this operation efficiently, a high level temporal representation is needed which allows the author to specify all the temporal dependencies between multimedia objects. In this paper, we propose an interval-based temporal model and constraints which provide a basis for the management of the consistency of multimedia documents. We propose an efficient algorithm allowing the detection of a wide range of inconsistencies. The emphasis in the design of these algorithms is put on the handling of both the flexibility of temporal specifications and the indeterministic behaviour of some media objects. Furthermore, we use the logical organization of the document in nested entities to enhance the performance of the methods used for detecting inconsistencies. The aim of our approach is to fulfill the following requirements . Structured modelling: The document content is defined by a hierarchy of nested components where leaves are basic media objects and nodes composite objects. . Incremental manipulation of the document: This means that the author adds one constraint at a time, retaining all previously introduced relations as the current state of the document. • Consistency checking: For every specification introduced by the author, the system checks if there exists a solution to the set of constraints, otherwise an error is reported to the author. • Integration of the different dimensions of multimedia documents: In our model, we focus on four dimensions of the document (logical, temporal, spatial and hyperlink). These dimensions are closely related and have to be considered simultaneously as they may interfere. In this paper, we will focus on the temporal and logical dimensions of the document but we will describe briefly how this fits with the spatial one. . Automatic production of temporal layout: The author specifies relations using a high-level language to describe the document temporal layout, and does not have to explicitly set the low-level synchronization dates. In section 2, we present our model of document composition showing the logical, spatial, and temporal dimensions. In section 3, we describe how we can create and manipulate the internal representation of the multimedia document in the form of constraint networks. In section 4, a consistency checking algorithm is proposed, followed by a discussion. Finally, in section 5, we conclude our work and give some future directions.

While a lot of work has been done in real-time systems, in real-time database systems, and more recently, in real-time active databases on the topic of time constrained processing of tasks and transactions, very little work exists that deals with the origin of the time constraints associated with the data, the events, and the actions. In this paper we attempt to identify the sources and semantics of time constraints and show that it is important to minimize the number of “mandated” timing requirements and also weaken the implications of timing constraint violations. The timing properties of interest and also the actions to be taken when the properties are violated can be specified using the Event-Condition-Action rules of active real-time databases. That is, an active real-time database can be made to store the data pertaining to the controlled system as well as the meta-data about the controlling system.

We present a system that can determine temporal scheduling of actions of an autonomous mobile robot by constructing a plan taking into account temporal constraints. Both qualitative and metric temporal information are represented in a constraint based network. Planning tasks are performed on the basis of the knowledge derived by this temporal knowledge representation system. Ordering temporally qualified actions falls out from maintaining temporal consistency. We applied it to plan under temporal constraints the actions of a manipulator mounted on a mobile platform in the case of the maintenance intervention over a hydraulic circuit.

Although a computer system cannot ensure that a user will act in a timely manner, it is possible to build systems that will guarantee that information is presented at the correct time, and will act upon user directions within an imposed deadline. A real-time database model is presented that is used to structure all interaction between users and the real-time system. Such a database has absolute and relative temporal consistencies defined. From such a model it is possible to generate the timing requirements for a set of tasks that can be guaranteed to retain temporal consistency in the database. It is shown how the database model can be used to define an appropriate architecture for HCI in this application domain.

In grid workflow systems, a checkpoint selection strategy is responsible for selecting checkpoints for conducting temporal verification at the runtime execution stage. Existing representative checkpoint selection strategies often select some unnecessary checkpoints and omit some necessary ones because they cannot adapt to the dynamics and uncertainty of runtime activity completion duration. In this article, based on the dynamics and uncertainty of runtime activity completion duration, we develop a novel checkpoint selection strategy that can adaptively select not only necessary, but also sufficient checkpoints. Specifically, we introduce a new concept of minimum time redundancy as a key reference parameter for checkpoint selection. An important feature of minimum time redundancy is that it can adapt to the dynamics and uncertainty of runtime activity completion duration. We develop a method on how to achieve minimum time redundancy dynamically along grid workflow execution and investigate its relationships with temporal consistency. Based on the method and the relationships, we present our strategy and rigorously prove its necessity and sufficiency. The simulation evaluation further demonstrates experimentally such necessity and sufficiency and its significant improvement on checkpoint selection over other representative strategies.

Songs sung by male humpback whales consist of distinct, pulsed sounds that are designated as units. Units are produced in some sequence to form a phrase, a repeated set of phrases forms a theme, and repeated themes form a song. A song can last from minutes to hours. The songs of eight humpback whales were recorded with a vertical array of five hydrophones spaced 7 m apart with the array located within 100 m of the whales. At least seven distinct units were identified aurally from this data set obtained during the 2002 winter humpback whale session in Hawaii. Four distinct recurring themes were found in the songs, and for each whale at least two themes were recorded. The average duration of each unit sampled and the silent interval following the unit were determined in order to describe the temporal characteristics of the songs. From the data the temporal consistency and cadence control of unit production by each humpback whale and between whales were determined. Understanding the temporal and spectral characteristics of units within songs and how these units vary between whales could ultimately help in the design of computer algorithms to automatically identify individual whales.

An item‐based survey was developed to evaluate how effectively physically‐based models can yield sounds that appear realistic to the ear. Based on past research and current insights, realism as a construct was hypothesized to comprise eight orthogonal factors: detail, physical plausibility (biological and nonbiological), temporal consistency, vividness, presence, and whether a sound evokes sensory images and associated memories. Sets of 8–9 declarative statements were selected by committee for each factor from large initial lists of such statements. Another committee selected weak, moderate and strong sound exemplars for each factor from digital libraries that were felt to vary uniquely along that factor. In an exploratory phase, 82 participants rated how representative each sound exemplar was of each statement. Based on factor analyses, the three highest loading statements for each factor were compared across sound exemplars; statistically significant differences were found between exemplars for all factors. A subsequent confirmatory phase used new sounds, coupled with synthetic versions from physically‐based simulations, to validate the proposed realism factors. Data from an additional 200 participants ratings for these stimuli supported the surveys construct validity. The survey should benefit sound engineers wishing to evaluate the degree to which their synthesis algorithms generate realistic output.

Snow avalanches in Cascade Canyon, Grand Teton National Park, have a significant effect on the region's vegetation mosaic, increasing the area's community diversity and creating a fragmented vegetation pattern. The structure and persistence of communities within avalanche tracks is a function of the frequency of avalanche occurrence. In many areas, shrub cover and conifer density increase as avalanche frequency increases. Conifers decrease in size yet increase in age as avalanches occur more often, a consequence of slower growth rates in trees within avalanche tracks. Because the probability of avalanche damage to conifers is related to the size of the tree, slow growth rates result in small trees that can survive many years in avalanche tracks, contributing to the persistence of the avalanche community. The canyon's vegetation patterns appear fairly stable due to the spatial and temporal consistency of avalanche occurrence and the persistence of communities in avalanche tracks. The primary effect avalanches have on this landscape is to increase the fragmentation of the vegetation patterns rather than to drive dynamic changes in the landscape mosaic.

Fish were sampled monthly at seven sites for 1 yr in Sister Grove Creek, an intermittent prairie stream in N-central Texas. Despite drastic seasonal fluctuations in discharge, abundance of individual species varied more spatially among sites than temporally at individual sites. Spatial differences in assemblage structure were attributed to differences in habitat and to migration from a downstream reservoir. The preponderance of spatial variation in species abundances combined with minimal temporal variation suggests: (1) temporal consistency of species abundance throughout the year, and (2) the highly dynamic hydrologic regime may not be a disturbance for Sister Grove Creek fishes.

The research was designed to investigate the role in children's trust in peers of (1) mutuality, which is the correspondence between children's trust in each other; (2) temporal consistency, which is the consistency of behavior across time and situations; and (3) helpfulness to peers. Kindergarten, second-grade, and fourth-grade children were tested. The children judged how much they trusted their peers and hypothetical peers who were either temporally consistent or temporally inconsistent. The teachers of the children rated each one on temporal consistency and on helpfulness to peers. As evidence for mutuality, it was found that (1) children's perceptions of each other's trustworthiness were correlated, and (2) a correlation was evident in which the less the children were trusted by peers, the less they trusted peers. Mixed support was provided for the hypothesized age shift on the basis of children's peer trust, from recent helpfulness in young children to temporal consistency in older children. As expe...

Although unrest is becoming increasingly prevalent in South African townships, there is a paucity of psychological research on the subject. The present article outlines the development of the 42-item Township Life Events Scale (TLES), a measure designed to assess the incidence and impact of township unrest. Internal and temporal consistency are satisfactory. The TLES also shows acceptable convergent and divergent validity. Future research priorities for the TLES and the impact of unrest on workplace behaviour in general are identified.

What are the factors responsible for skilled typing performance, and do they change with the age of the typist? These questions were addressed in two studies by examining time and accuracy of keystrokes in a variety of typinglike activities among typists ranging in speed from 17 to 104 net words per minute and ranging in age from 19 to 72 years old. Typing skill was related to the temporal consistency of making the same keystroke, the efficiency of overlapping successive keystrokes, the speed of alternate-hand tapping, and the number of characters of to-be-typed text required to maintain a normal rate of typing. Older typists were slower in tapping rate and in choice reaction time but were not slower in speed of typing, apparently because they were more sensitive to characters farther in advance of the currently typed character than young typists.

Spatial variations in the price of chemists' goods in Dublin are examined for a number of different baskets of products and for two time periods, so as to tackle the problem of the ‘sample of one’. A number of hypotheses including measures of store size, proximity of competition, and the type of residential area in which the shop is located are examined by means of bivariate and multiple correlation in relation to price variations, and in general the results are consistent for different baskets at a single time period: shops which are expensive for one combination of products tend to be expensive for most product combinations. However, there is little consistency of results between the two time periods, and this must create problems for price-motivated consumers: an inexpensive shop at one time period may not necessarily be inexpensive at a subsequent time period even for the same products. Even so, there is some temporal consistency in that the size of shop generally has a direct relationship with variations in prices, smaller shops being the more expensive.

This study explores the allocation mechanisms in the public provision of interregional transport and communication infrastructure in Sweden. The logistics of the flow of infrastructure investments from the government budget is examined. The purpose is to consider how the demand for infrastructure investment is estimated by public agencies, how judgments are made on governmental expenditure policy, and what determines the spatial distribution of infrastructure investment. The analysis is carried out along four main lines; demand for infrastructure services, political intentions and goals, the modal composition of investments for transport and communication, and the spatial distribution of infrastructure. The discussion considers functional, spatial, and temporal consistency aspects in the public provision of infrastructure.

The present study used a prospective design to assess the ability of the reformulated learned helplessness model of depression to predict affective reactions to a stressful event. A total of 131 university students completed measures of depressive attributional style, trait depression, and performance aspirations for an upcoming course examination. Immediately following the exam, subjects completed a state depression measure; subjects again reported their state depression immediately following receipt of their examination grades. The discrepancy between subjects' grade aspirations and their achieved grades was used as a measure of subjective stress. Analyses indicated that a depressive attributional style interacted with this measure of subjective stress to predict changes in depressive mood following the examination, but not after receipt of their grades. The pattern of results is discussed in light of (a) the need for research investigating temporal factors in vulnerabilities for depressive mood and (b) the limitations of the study.

An initial sample of n = 316 MMPI profiles, produced by inmates at a Security Level III Federal Correctional Institution in Ashland, Kentucky, were subjected to the classification rules delineated by Megargee and Bohn (1979). Only 22 of these 316 profiles failed to meet the criteria for inclusion in at least one of the 10 Myer-Megargee types. The resultant 294 classifiable profiles (93%) were uniquely classified using only minimal (Set I) rules in combination with the accessory (Set II) rules provided by Megargee and Bohn (1979). Such a uniquely classifiable sample was intentionally selected for a potential bias towards stability. A final sample of n = 85 inmates were voluntarily retested (median test-retest interval = 9.78 months) and were classified. Only 16 of 85 inmates retained their original type designation upon retesting. On a subset of the overall sample (median test-retest = 3.15 months) only one of 14 profiles remained unchanged. These results both parallel and extend the data presented by Simmons, Johnson, Gouvier, and Muzyczka (1981) indicating marked instability sf the Myer-Megargee inmate typology. Thus, further data are amassed that question the efficacy of the Myer-Megargee typology for correctional decision-making.

One hundred nonpatient adults, screened for evidence of personality disorganization, were retested after a three-year interval to study the temporal consistency of the Rorschach. In general, the correlational analysis for 19 basic variables studied, and a directionality analysis for five ratios, illustrate a considerable sturdiness over time. It is postulated that these variables can be separated into situational related indices (state variables), and more stable scores indicative of durable response styles (trait variables), based on the assumption that variables with lower correlations would identify state variables, while the higher correlations would signify the trait features. Nine of the 19 variables yield retest correlations in excess of.

Summary Regular drinkers identified in a general population survey of adults in the Lothian region were followed up 3 or 4 years later, with respect to alcohol-related problems. Using a variety of analytic methods, a moderate degree of temporal consistency in problems was found, for total number of problems, and for individual problems, with feeling ashamed of having done something while drunk being the most consistent.

• Symptoms of psychosis and psychoticlike phenomena were systematically inquired for, using a semistructured diagnostic interview protocol, in 58 prepubertal children who fulfilled Research Diagnostic Criteria for major depressive disorder. Subjects were classified into endogenous and nonendogenous subtypes. Forty-eight percent of the sample reported hallucinations. Thirty-six percent (including 48% of those in the endogenous subtype and 24% of those in the nonendogenous subtype) reported auditory hallucinations that consisted of one or more words other than their names, experienced in clear consciousness. Sixteen percent reported visual or tactile hallucinations. Four children were rated as having delusional ideas. Symptoms were analyzed in terms of content, formal characteristics, thematic and temporal consistency with depressed mood, extent of the child's belief in the reality of the experience, frequency, and experienced location of auditory hallucinations. The psychopathologic meaning of the reported phenomena is uncertain.

The temporal consistency for 23 Rorschach variables was examined in an 8-year longitudinal study which involved testing 57 subjects at 24-month intervals, beginning when the subjects were age 8 and administering the fourth retest when they were age 16. The objective was to compare these data with norms for the various age groups and to study the features of development as manifest in the Rorschach. Only one variable, the X+%, appears relatively stable from age 8 to age 16, and the majority do not manifest consistency until the interval between ages 14 and 16. The data suggest that long term descriptive predictions of children can only be formulated with great caution except when conditions of severe pathology exist.

A quantitative approach to characterization of unit response in the visual system was proposed. Temporal parameters of response were analyzed by post stimulus time-histograms obtained from point visual stimulation. Explicit temporal intervals were chosen to coincide with discrete response components. By calculating net firing rate within these intervals as a function of stimulus position, quantitative receptive field maps were constructed. To estimate response between stimulated matrix mesh points, a general contouring program was described. Latency, and synchronicity — a logarithmic probability function sensitive to temporal consistency of response — were also mapped as functions of stimulus position. Representative firing rate maps morphologically resembled previously described receptive fields; latency and synchronicity maps, it is argued, allow for additional mechanistic and functional analysis of striate units. By explicitly separating temporal and spatial variables and quantifying unit response, the present methods will hopefully enlarge the scope and sensitivity of receptive field analysis.

Representing and reconstructing 3D deformable shapes are two tightly linked problems that have long been studied within the computer vision field. Deformable shapes are truly ubiquitous in the real world, whether be it specific object classes such as humans, garments and animals or more abstract ones such as generic materials deforming under an external force. Practical computer vision algorithms must be able to understand the shapes of objects in the observed scenes to unlock the wide spectrum of much sought after applications ranging from virtual try-on to automated surgeries. Automatic shape reconstruction is known to be an ill-posed problem, especially in the common scenario of a single image input. Therefore, the modern approaches rely on deep learning paradigm which has proven to be extremely effective even for the severely under-constrained computer vision problems. We, too, exploit the success of data-driven approaches, however, we also show that generic deep learning models can greatly benefit from being combined with explicit knowledge originating in computational geometry. We analyze the use of various 3D shape representations and we distinctly focus on one of them, the atlas-based representation, which turns out to be especially suitable for modeling deformable shapes and which we further improve and extend. The atlas-based representation models the surfaces as an ensemble of continuous functions and thus allows for arbitrary resolution and analytical surface analysis. We identify its major shortcomings, namely the patch collapse, patch overlap and strong mapping distortions, and we propose novel regularizers based on analytically computed properties of the reconstructed surfaces. Our approach counteracts the aforementioned drawbacks while yielding higher reconstruction accuracy. We dive into the problematics of atlas-based shape representation deeper and focus on another design flaw, the global inconsistency of the mappings. While it is not reflected in quantitative metrics, it is detrimental to the visual quality of the reconstructed surfaces. Specifically, we design loss functions encouraging intercommunication among the mappings which pushes the resulting surface towards a C1 smooth function and thus dramatically improves the visual quality. Furthermore, we adapt the atlas-based representation so that it could model a full sequence of a deforming object in a temporally-consistent way. The goal is to produce such reconstruction where each surface point always represents the same semantic point on the target GT surface. To achieve such behavior, we note that if each surface point deforms close-to-isometrically, its semantic location likely remains unchanged. Practically, we make use of the Riemannian metric, and force it to remain point-wise constant throughout the sequence. The experiments show that our method yields SotA results on correspondence estimation task. Finally, we look into a particular problem of monocular texture-less deformable shape reconstruction. We propose a multi-task learning approach which jointly produces a normal map, a depth map and a mesh corresponding to the observed surface. We show that producing multiple different 3D representations of the same objects results in higher reconstruction quality. We acquire a large real-world annotated dataset of texture-less deforming objects and we release it for public use.

We propose a novel approach to reconstructing curvilinear tree structures evolving over time, such as road networks in 2D aerial images or neural structures in 3D microscopy stacks acquired in vivo . To enforce temporal consistency, we simultaneously process all images in a sequence, as opposed to reconstructing structures of interest in each image independently. We formulate the problem as a Quadratic Mixed Integer Program and demonstrate the additional robustness that comes from using all available visual clues at once, instead of working frame by frame. Furthermore, when the linear structures undergo local changes over time, our approach automatically detects them.

Modern methods for counting people in crowded scenes rely on deep networks to estimate people densities in individual images. As such, only very few take advantage of temporal consistency in video sequences, and those that do only impose weak smoothness constraints across consecutive frames. In this paper, we advocate estimating people flows across image locations between consecutive images and inferring the people densities from these flows instead of directly regressing. This enables us to impose much stronger constraints encoding the conservation of the number of people. As a result, it significantly boosts performance without requiring a more complex architecture. Furthermore, it also enables us to exploit the correlation between people flow and optical flow to further improve the results. We will demonstrate that we consistently outperform state-of-the-art methods on five benchmark datasets., Comment: ECCV 2020

This paper presents a new approach for estimating the disparity search range in stereo video that enforces temporal consistency. Reliable search range estimation is very important since an incorrect estimate causes most stereo matching methods to get trapped in local minima or produce unstable results over time. In this work, the search range is estimated based on a disparity histogram that is generated with sparse feature matching algorithms such as SURF. To achieve more stable results over time, we further propose to enforce temporal consistency by calculating a weighted sum of temporally- neighboring histograms, where the weights are determined by the similarity of depth distribution between frames. Experimental results show that this proposed method yields accurate disparity search ranges for several challenging stereo videos and is robust to various forms of noise, scene complexity and camera configurations.

As in other application areas, there is an increasing need for managing a large amount of data in the real-time area. This need gave birth to a system called a real-time database system (RTDBS) that provides database operations with timing constraints. One typical timing constraint in an RTDBS is temporal consistency that states that a transaction must read temporally valid data objects. This requires the data objects to be updated repeatedly. This paper proposes three novel schemes that aim to minimize the number of updates of data objects needed for guaranteeing the temporal consistency requirements of transactions with hard-deadlines. The three guarantee schemes differ from each other in the amount of updates and the implementation complexity. This paper also gives a framework for integrating transactions with soft-deadlines into the three guarantee schemes.

The highly specialized nature of medical domains has increased the demand of consultation of medical specialists to solve certain therapeutic and diagnostic problems. This problem requires the development of complex Knowledge Acquisition tools to build this Knowledge Base (KB). In this work, we present a KB System for medical domains which has shown the benefits of approaching the KB design from a wider perspective, allowing users to provide, query and browse knowledge.

The main aim was to study the predictive power of sense of coherence regarding future needs of care among elderly patients evaluated as medically ready for discharge from somatic emergency care. A secondary aim was to study the consistency of sense of coherence over time among patients with this kind of experience. The sample consisted of 53 Swedish patients (mean age 82.8 years, SD = 6.6 years) who had completed their medical treatment at surgical or orthopaedic departments. The predominant diagnosis was lower limb fractures. Sense of coherence was assessed twice, on the day the patient was evaluated as medically ready for discharge and 1 month later. On the second assessment occasion, 28 patients had returned to their homes, 17 were staying at institutions, and eight had died. Patients who returned to their homes reported the strongest sense of coherence while still in hospital. Patients who were staying at institutions scored lowest on the overall sense of coherence scale and on the comprehensibility subscale. Patients who died before the second measurement occasion scored lowest on the meaningfulness subscale. A correlation of 0.51 was noted between the two assessments of sense of coherence, indicating a moderate temporal consistency.

In semi-natural environments sows often select a protected nest site at farrowing. Investigations of nest site selection under indoor conditions are scarce but suggest that sows prefer to nest and farrow under a covered area. Since feed-back from a functional nest may improve maternal behaviour and thus piglet survival we investigated if access to a covered area and ad libitum access to straw would improve maternal behaviour of importance for piglet survival.We used forty-four primiparous sows housed in large individual pens from approximately 2 weeks prior to expected parturition. The sows were randomly distributed to four treatment groups of +/- ad libitum access to straw combined with +/- access to a covered area in the pen. The sows' position in the pen (left or right side) was observed from 24 h prior to birth of the first piglet until 24 h after using 10-min interval scan sampling and the temporal pattern of sow position in relation to position of the cover (left or right side of the pen) was studied. Furthermore, postural and nest-building behaviours were observed, as well as situations where the piglets were either at risk of being crushed or were crushed.The results showed no significant influence of cover on sow position. Odds-ratio for choosing the covered farrowing position was 1.4 (95% creditability interval (0.4-5.1)). The temporal pattern of sow position showed that the probability of observing the sows in the part of the pen classified as the farrowing position increased from close to 0.5 to 1 as parturition approached and stayed high during the 24 h observation period after birth of the first piglet, except for a small drop around 12 h after birth of the first piglet. However, the tendency to stay in the nest during farrowing was not increased neither by cover nor by ad libitum access to straw. Access to cover significantly increased prepartum nest-building behaviour (F1,29 = 4.27, P = 0.05) whereas there was no significant influence of access to straw. However, it should be taken into account that all sows had access to sawdust on the floor, which may also serve as a nesting material.The number of crushing and near crushing situations within 24 h after birth of first piglet (BFP) was significantly decreased in pens with access to straw (F1,25 = 4.50, P = 0.04), whereas there was no significant influence of cover on this variable. Neither the number of scan intervals with lateral lying nor the number of postural changes were significantly influenced by cover or straw.In conclusion, the sows did select a specific nest site within the farrowing pen, but neither the choice of nest site nor the temporal consistency of staying in the nest were significantly affected by cover or ad libitum access to straw. Access to a covered area increased the occurrence of nest-building behaviour whereas access to straw reduced the number of near crushing situations.

Modern methods for counting people in crowded scenes rely on deep networks to estimate people densities in individual images. As such, only very few take advantage of temporal consistency in video sequences, and those that do only impose weak smoothness constraints across consecutive frames. In this paper, we advocate estimating people flows across image locations between consecutive images and inferring the people densities from these flows instead of directly regressing them. This enables us to impose much stronger constraints encoding the conservation of the number of people. As a result, it significantly boosts performance without requiring a more complex architecture. Furthermore, it allows us to exploit the correlation between people flow and optical flow to further improve the results. We also show that leveraging people conservation constraints in both a spatial and temporal manner makes it possible to train a deep crowd counting model in an active learning setting with much fewer annotations. This significantly reduces the annotation cost while still leading to similar performance to the full supervision case., Comment: Accepted to IEEE Transactions on Pattern Analysis and Machine Intelligence. Extension of Our ECCV 2020 Paper: arXiv:1911.10782

Different ensemble-based data assimilation (DA) approaches for palaeoclimate reconstructions have been recently undertaken, but no systematic comparison among them has been attempted. We compare an off-line and an on-line ensemble-based method, with the testing period being the 17th century, which led into the Maunder Minimum. We use a low-resolution version of Max Planck Institute for Meteorology Earth System Model (MPI-ESM) to assimilate the Past Global Changes (PAGES) 2k continental temperature reconstructions. In the off-line approach, the ensemble for the entire simulation period is generated first and then the ensemble is used in combination with the empirical information to produce the analysis. In contrast, in the on-line approach, the ensembles are generated sequentially for sub-periods based on the analysis of previous sub-periods. Both schemes perform better than the simulations without DA. The on-line method would be expected to perform better if the assimilation led to states of the slow components of the climate system that are close to reality and the system had sufficient memory to propagate this information forward in time. In our comparison, which is based on analysing correlations and differences between the analysis and the proxy-based reconstructions, we find similar skill for both methods on the continental and hemispheric scales. This indicates either a lack of control of the slow components in our setup or a lack of skill in the information propagation on decadal timescales. Additional experiments are however needed to check whether the conclusions reached in this particular setup are valid in other cases. Although the performance of the two schemes is similar and the on-line method is more difficult to implement, the temporal consistency of the analysis in the on-line method makes it in general preferable.

The nature of the Rorschach is discussed using data from three temporal consistency studies with children of different ages, plus two other studies in which the retest reliability question was approached by experimental manipulation. It is postulated that the test response reflects the basic psychological styles or response tendencies that have developed within the personality, plus representations of the more private world of the person. The challenge for interpretation and research is to differentiate which of the test elements reflect each of these features, and to determine how they merge to create the response during a very brief time interval.