scholarly journals Finding Action Tubes with a Sparse-to-Dense Framework

2020 ◽  
Vol 34 (07) ◽  
pp. 11466-11473
Author(s):  
Yuxi Li ◽  
Weiyao Lin ◽  
Tao Wang ◽  
John See ◽  
Rui Qian ◽  
...  
Keyword(s):  
Dynamic Feature ◽  
Video Streams ◽  
Short Term ◽  
Key Characteristics ◽  
Benchmark Datasets ◽  
Tube Output ◽  
Spatio Temporal ◽  
First Time ◽  
Temporal Action

The task of spatial-temporal action detection has attracted increasing researchers. Existing dominant methods solve this problem by relying on short-term information and dense serial-wise detection on each individual frames or clips. Despite their effectiveness, these methods showed inadequate use of long-term information and are prone to inefficiency. In this paper, we propose for the first time, an efficient framework that generates action tube proposals from video streams with a single forward pass in a sparse-to-dense manner. There are two key characteristics in this framework: (1) Both long-term and short-term sampled information are explicitly utilized in our spatio-temporal network, (2) A new dynamic feature sampling module (DTS) is designed to effectively approximate the tube output while keeping the system tractable. We evaluate the efficacy of our model on the UCF101-24, JHMDB-21 and UCFSports benchmark datasets, achieving promising results that are competitive to state-of-the-art methods. The proposed sparse-to-dense strategy rendered our framework about 7.6 times more efficient than the nearest competitor.

Collaborative Local-Global Learning for Temporal Action Proposal

2021 ◽  
Vol 12 (5) ◽  
pp. 1-14
Author(s):  
Yisheng Zhu ◽  
Hu Han ◽  
Guangcan Liu ◽  
Qingshan Liu
Keyword(s):  
Short Term Memory ◽  
Background Suppression ◽  
Global Learning ◽  
Short Term ◽  
Term Memory ◽  
Temporal Intervals ◽  
Benchmark Datasets ◽  
Long Short Term Memory ◽  
Temporal Action

Temporal action proposal generation is an essential and challenging task in video understanding, which aims to locate the temporal intervals that likely contain the actions of interest. Although great progress has been made, the problem is still far from being well solved. In particular, prevalent methods can handle well only the local dependencies (i.e., short-term dependencies) among adjacent frames but are generally powerless in dealing with the global dependencies (i.e., long-term dependencies) between distant frames. To tackle this issue, we propose CLGNet, a novel Collaborative Local-Global Learning Network for temporal action proposal. The majority of CLGNet is an integration of Temporal Convolution Network and Bidirectional Long Short-Term Memory, in which Temporal Convolution Network is responsible for local dependencies while Bidirectional Long Short-Term Memory takes charge of handling the global dependencies. Furthermore, an attention mechanism called the background suppression module is designed to guide our model to focus more on the actions. Extensive experiments on two benchmark datasets, THUMOS’14 and ActivityNet-1.3, show that the proposed method can outperform state-of-the-art methods, demonstrating the strong capability of modeling the actions with varying temporal durations.

scholarly journals Spatio-temporal associations of air pollutant concentrations, GP respiratory consultations and respiratory inhaler prescriptions: a 5-year study of primary care in the borough of Lambeth, South London

2021 ◽  
Vol 20 (1) ◽  
Author(s):  
Mark Ashworth ◽  
◽  
Antonis Analitis ◽  
David Whitney ◽  
Evangelia Samoli ◽  
...  
Keyword(s):  
Primary Care ◽  
Negative Binomial ◽  
Hospital Admissions ◽  
Air Pollutant ◽  
Respiratory Morbidity ◽  
Short Term ◽  
Short Term Exposure ◽  
Short And Long Term ◽  
Spatio Temporal

Abstract Background Although the associations of outdoor air pollution exposure with mortality and hospital admissions are well established, few previous studies have reported on primary care clinical and prescribing data. We assessed the associations of short and long-term pollutant exposures with General Practitioner respiratory consultations and inhaler prescriptions. Methods Daily primary care data, for 2009–2013, were obtained from Lambeth DataNet (LDN), an anonymised dataset containing coded data from all patients (1.2 million) registered at general practices in Lambeth, an inner-city south London borough. Counts of respiratory consultations and inhaler prescriptions by day and Lower Super Output Area (LSOA) of residence were constructed. We developed models for predicting daily PM2.5, PM10, NO2 and O3 per LSOA. We used spatio-temporal mixed effects zero inflated negative binomial models to investigate the simultaneous short- and long-term effects of exposure to pollutants on the number of events. Results The mean concentrations of NO2, PM10, PM2.5 and O3 over the study period were 50.7, 21.2, 15.6, and 49.9 μg/m3 respectively, with all pollutants except NO2 having much larger temporal rather than spatial variability. Following short-term exposure increases to PM10, NO2 and PM2.5 the number of consultations and inhaler prescriptions were found to increase, especially for PM10 exposure in children which was associated with increases in daily respiratory consultations of 3.4% and inhaler prescriptions of 0.8%, per PM10 interquartile range (IQR) increase. Associations further increased after adjustment for weekly average exposures, rising to 6.1 and 1.2%, respectively, for weekly average PM10 exposure. In contrast, a short-term increase in O3 exposure was associated with decreased number of respiratory consultations. No association was found between long-term exposures to PM10, PM2.5 and NO2 and number of respiratory consultations. Long-term exposure to NO2 was associated with an increase (8%) in preventer inhaler prescriptions only. Conclusions We found increases in the daily number of GP respiratory consultations and inhaler prescriptions following short-term increases in exposure to NO2, PM10 and PM2.5. These associations are more pronounced in children and persist for at least a week. The association with long term exposure to NO2 and preventer inhaler prescriptions indicates likely increased chronic respiratory morbidity.

scholarly journals A SLR Pre-Processing Algorithm Based on Satellite Signature Effect

2021 ◽  
Vol 11 (19) ◽  
pp. 8880
Author(s):  
Bowen Guan ◽  
Cunbo Fan ◽  
Ning An ◽  
Ricardo Cesar Podesta ◽  
Dra Ana Pacheco ◽  
...  
Keyword(s):  
Mean Value ◽  
Observation Data ◽  
Short Term ◽  
Error Sources ◽  
Effect Model ◽  
The Mean ◽  
The Stability ◽  
First Time ◽  
Satellite Signature Effect

As one of the major error sources, satellite signature effect should be reduced or even erased from the distribution of the post-fit residuals to improve the ranging precision. A simulation of satellite signature effect removal process for normal point algorithm is conducted based on a revised model of satellite response, which fully considers the structural and distribution characteristics of retroreflectors. In order to eliminate both long-term and short-term satellite signature effect, a clipping method for SLR data processing is proposed by defining the clipping location as 5.6 mm away from the mean value of the long-term fit residuals to select effective returns for normal points. The results indicate that, compared to normal points algorithm, the RMS per NP of LAGEOS-1 observation data processed by the clipping method is reduced from 62.90 ± 9.9 mm to 56.07 ± 4.69 mm, and the stability of RMS is improved 53%. This study improves the satellite signature effect model and simulates the fluctuation of normal points caused by satellite signature effect for the first time. The new method based on the simulation of satellite signature effect has stronger robustness and applicability, which can further minimize the influence of satellite signature effect on the SLR production and significantly improve the data property.

scholarly journals Assessing the Spatial and Spatio-Temporal Distribution of Forest Species via Bayesian Hierarchical Modeling

Forests ◽  
2018 ◽  
Vol 9 (9) ◽  
pp. 573 ◽  
Author(s):  
Óscar Rodríguez de Rivera ◽  
Antonio López-Quílez ◽  
Marta Blangiardo
Keyword(s):  
Hierarchical Modeling ◽  
Temporal Distribution ◽  
Short Term ◽  
Distribution Models ◽  
Address Climate Change ◽  
Bayesian Hierarchical ◽  
Temporal Models ◽  
Spatio Temporal ◽  
The Relationship

Climatic change is expected to affect forest development in the short term, as well as the spatial distribution of species in the long term. Species distribution models are potentially useful tools for guiding species choices in reforestation and forest management prescriptions to address climate change. The aim of this study is to build spatial and spatio-temporal models to predict the distribution of four different species present in the Spanish Forest Inventory. We have compared the different models and showed how accounting for dependencies in space and time affect the relationship between species and environmental variables.

Foreign Demand, Investment And Trade Balance: The Case Of Australia

2011 ◽  
Vol 1 (4) ◽  
Author(s):  
Jean-Francois Hoarau ◽  
Alain Nurbel ◽  
Nelson Latchimy
Keyword(s):  
Trade Balance ◽  
High Growth ◽  
Short Term ◽  
Foreign Demand ◽  
J Curve ◽  
The World ◽  
Curve Effect ◽  
First Time ◽  
The Empirical Analysis

This paper aims at analysing the relation between real trade balance and foreign demand in the case of a small opened economy, which highly depends upon the rest of the world for productive capital. Theoretical analysis allows us to bring forth a kind of “J-curve” effect. Indeed, when foreign demand for domestic goods increases, the country is to import in a first time in order to improve its productive capacities, resulting in worsening trade balance. However, in a second time, once the cumulated capital inventory became sufficient, the trade balance improves under the pressure of domestic exports high growth. The empirical analysis based on Australia from 1982 (1) to 2001 (1) supports this theory. We show there are negative short term and positive long term elasticities.

scholarly journals Building bridges between experts and the public: a comparison of two-way communication formats for flooding and air pollution risk

2018 ◽  
Author(s):  
Maria Loroño-Leturiondo ◽  
Paul O'Hare ◽  
Simon J. Cook ◽  
Stephen R. Hoon ◽  
Sam Illingworth
Keyword(s):  
Air Pollution ◽  
Serious Games ◽  
Short Term ◽  
Pollution Risk ◽  
The Public ◽  
Safe Spaces ◽  
Personal Experiences ◽  
Educational Programmes ◽  
First Time

Abstract. Urban centres worldwide are adversely affected by flooding and air pollution. Better prepared citizens are crucial to limiting the impacts of these hazards, and both lay knowledge and personal experiences are important in complementing and challenging expert opinion. For the first time, this study offers a critical comparison of how different two-way communication formats have been used worldwide between experts and the public in relation to flooding and air pollution risk. Through a systematic review, we analyse social media, educational programmes, serious games, citizen science, and forums in terms of their effectiveness in respect of communicating short-term incidents, long-term awareness, and long-term knowledge in the context of flooding and air pollution risk. We find that there is neither a one-size-fits-all, nor superior, format of communication. No single format is effective in fulfilling all three communication purposes. All five formats analysed appear to be successful under different circumstances and are never representative of all segments of the population. Communication between experts and the public is difficult and full of tensions, information alone is not enough. Our study shows different ways of incorporating strategies to build trust between experts and the public and make communication more fun and accessible, breaking down hierarchies and creating safe spaces for co-creation where everyone feels empowered to participate and benefits.

scholarly journals Complex networks and waveforms from acoustic emissions in laboratory earthquakes

2014 ◽  
Vol 21 (4) ◽  
pp. 763-775 ◽  
Author(s):  
H. O. Ghaffari ◽  
B. D. Thompson ◽  
R. P. Young
Keyword(s):  
Acoustic Emission ◽  
Complex Networks ◽  
Acoustic Emissions ◽  
Dissipated Energy ◽  
Short Term ◽  
Thermally Activated ◽  
Direct Observations ◽  
Functional Complex ◽  
First Time

Abstract. Understanding the physics of acoustic excitations emitted during the cracking of materials is one of the long-standing challenges for material scientists and geophysicists. In this study, we report novel results of applications of functional complex networks on acoustic emission waveforms emitted during the evolution of frictional interfaces. Our results show that laboratory faults at microscopic scales undergo a sequence of generic phases, including strengthening, weakening or fast slip and slow slip, leading to healing. For the first time we develop a formulation on the dissipated energy due to acoustic emission signals in terms of short-term and long-term features (i.e., networks' characteristics) of events. We illuminate the transition from regular to slow ruptures. We show that this transition can lead to the onset of the critical rupture class similar to the direct observations of this phenomenon in the transparent samples. Furthermore, we demonstrate the detailed submicron evolution of the interface due to the short-term evolution of the rupture tip. As another novel result, we find that the nucleation phase of most amplified events follows a nearly constant timescale, corresponding to the initial strengthening or locking of the interface. This likely indicates that a thermally activated process can play a crucial role near the moving crack tip.

scholarly journals Statistical behaviours of earthquake occurrences in the Central Anatolian Region of Turkey: region-time-magnitude analyses

Baltica ◽  
2021 ◽  
pp. 157-173
Author(s):  
Serkan Öztürk
Keyword(s):  
Fault Zone ◽  
Earthquake Hazard ◽  
B Value ◽  
Short Term ◽  
Small Magnitude ◽  
Strong Earthquakes ◽  
Recurrence Times ◽  
Z Value ◽  
Spatio Temporal

The main objective of this work is to make detailed region-time-magnitude analyses by describing the statistical behaviours of earthquakes in the Central Anatolian Region of Turkey. In this scope, several seismic and tectonic parameters such as Mcomp, b-value, Dc-value, Z-value, recurrence times and annual probabilities were evaluated. For the analyses, a homogeneous catalogue including 10,146 earthquakes with 1.0 ≤ Md ≤ 5.7 between 30 July 1975 and 29 December 2018 was used and spatio-temporal changes of earthquake behaviours were mapped for the beginning of 2019. Earthquake magnitudes varied from 1.9 to 3.0 on average, and hence Mcomp was considered to be 2.6. The b-value was calculated as 1.26 ± 0.07, and this relatively large value indicates that small-magnitude events are dominant. The Dc-value was computed as 1.31 ± 0.03. This small value means that distances between epicentres approach the diameter of the cluster, and seismic activity is more clustered at smaller scales or in larger regions. The spatio-temporal analyses of recurrence times suggest that the Central Anatolian Region has an intermediate/long-term earthquake hazard in comparison to occurrences of strong earthquakes in the short term. Several anomaly regions of a small b-value and a large Z-value were found in and around the Tuzgölü Fault Zone, Central Anatolian Fault Zone, Salanda fault and Niğde fault at the beginning of 2019. Thus, a combination of the regions with a lower b-value, a higher Z-value and also moderate recurrence times may give significant clues for the future possible earthquakes, and detected regions may be thought to be the most likely areas for strong/large events in the Central Anatolian Region.

scholarly journals Covid-19: Pandemic of Domestic Violence

2020 ◽  
Vol 9 (2) ◽  
pp. 111-136
Author(s):  
Tatiana Stoianova ◽  
Liudmyla Ostrovska ◽  
Grygorii Tripulskyir
Keyword(s):  
Domestic Violence ◽  
Well Being ◽  
Legal Regulation ◽  
Short Term ◽  
Procedural Law ◽  
Special Means ◽  
Legislative Amendments ◽  
Economic Nature ◽  
First Time

The article is devoted to the analysis of domestic violence in the context of Covid-19. The research is carried out for the first time in the focus of several sciences: psychology, sociology, and jurisprudence. To study the legal regulation of domestic violence, knowledge from different branches of law was used: international, criminal, administrative, and civil procedural law. Attention was paid to the historical retrospective—how the concept of domestic violence first appeared at the world level, and how it was differentiated and implemented in the national legislation of the participating countries. The problems of signing the Istanbul Agreement are highlighted. Special attention was paid to the current wave of domestic violence as a result of the Covid-19 pandemic. The prerequisites of a general psychological, social, and economic nature, their interdependence, and connection with the pandemic were investigated. The scale of the scourge of domestic violence in the context of a pandemic in different countries is indicated, and its short-term and long-term consequences for the well-being of the nation. The specific mechanisms for preventing family violence at three levels are considered: general criminogenic, a comprehensive mechanism for preventing violence at the level of interaction between the state and public organizations, and directly special means. The study concludes that Covid-19 pandemic has a direct impact on the exacerbation of domestic violence. The solutions are proposed, from legislative amendments to the redistribution of state and public forces to address the problem of domestic violence.

A Logic-Based Approach to Activity Recognition

2013 ◽  
pp. 1-13
Author(s):  
Alexander Artikis ◽  
Marek Sergot ◽  
Georgios Paliouras
Keyword(s):  
Activity Recognition ◽  
Human Activities ◽  
Video Content ◽  
Short Term ◽  
Event Calculus ◽  
Video Frames ◽  
Detailed Evaluation ◽  
Spatio Temporal ◽  
Complex Activities

The authors have been developing a system for recognising human activities given a symbolic representation of video content. The input of the system is a stream of time-stamped short-term activities detected on video frames. The output of the system is a set of recognised long-term activities, which are pre-defined spatio-temporal combinations of short-term activities. The constraints on the short-term activities that, if satisfied, lead to the recognition of a long-term activity, are expressed using a dialect of the Event Calculus. The authors illustrate the expressiveness of the dialect by showing the representation of several typical complex activities. Furthermore, they present a detailed evaluation of the system through experimentation on a benchmark dataset of surveillance videos.

Sign in / Sign up

Export Citation Format

Share Document