Analysing Quality of Resilience in Fish4Knowledge Video Analysis Workflows

Author(s):  
Gayathri Nadarajan ◽  
Cheng-Lin Yang ◽  
Yun-Heh Chen-Burger ◽  
Rafael Tolosana-Calasanz ◽  
Omer F. Rana
Keyword(s):  
2020 ◽  
Vol 71 (7) ◽  
pp. 868-880
Author(s):  
Nguyen Hong-Quan ◽  
Nguyen Thuy-Binh ◽  
Tran Duc-Long ◽  
Le Thi-Lan

Along with the strong development of camera networks, a video analysis system has been become more and more popular and has been applied in various practical applications. In this paper, we focus on person re-identification (person ReID) task that is a crucial step of video analysis systems. The purpose of person ReID is to associate multiple images of a given person when moving in a non-overlapping camera network. Many efforts have been made to person ReID. However, most of studies on person ReID only deal with well-alignment bounding boxes which are detected manually and considered as the perfect inputs for person ReID. In fact, when building a fully automated person ReID system the quality of the two previous steps that are person detection and tracking may have a strong effect on the person ReID performance. The contribution of this paper are two-folds. First, a unified framework for person ReID based on deep learning models is proposed. In this framework, the coupling of a deep neural network for person detection and a deep-learning-based tracking method is used. Besides, features extracted from an improved ResNet architecture are proposed for person representation to achieve a higher ReID accuracy. Second, our self-built dataset is introduced and employed for evaluation of all three steps in the fully automated person ReID framework.


2019 ◽  
pp. 027112141985714
Author(s):  
Christan Grygas Coogle ◽  
Sarah Nagro ◽  
Kelley Regan ◽  
Kristen Merrill O’Brien ◽  
Jennifer R. Ottley

We used a multiple-probe single-case research design to examine the effect of a professional development package that included real-time technology-enhanced performance-based feedback and video analysis on three preschool teachers’ use of naturalistic instruction targeting children’s communication and child responses. We also measured the quality of the teachers’ naturalistic instruction targeting children’s communication. The professional development package was effective in enhancing teachers’ use of naturalistic instruction targeting children’s communication. In addition, children responded and teachers enhanced the quality of naturalistic instruction targeting children’s communication. Teachers maintained their use of naturalistic instruction targeting children’s communication upon removal of the intervention condition.


2020 ◽  
pp. 187-196

This study refers to the evolution of the Romanian Women’s Epee Team at the Olympic Games in terms of performance, but especially from the perspective of making up the team and establishing the strategy for addressing each goal-focused competition. The literature review reveals that one of the ways to increase both the quality of the training process in fencing and the success rate in team events lies in the team composition, but also in establishing the sequence of athletes competing in a match. Research purpose: The study aims to highlight the dynamics of the Romanian Women’s Epee Team through both a detailed analysis of the results achieved by athletes in their fight against various opponents and their comparison at different times of the matches. Methods: Bibliographic study, comparison, abstraction, induction and deduction methods, video analysis, mathematical and statistical methods. Results and conclusion: The sequence of athletes’ performance in team competitions and their effectiveness depending on the number of won/lost bouts, but also successful hits and recovery times were analysed based on the competition protocols for the Rio de Janeiro and London Olympics, and the performance of the Romanian Women’s Epee Team was compared with the performance achieved by other competing teams. The study has shown that, in today’s modern fencing (particularly the Epee), one of the most important factors for achieving performance in the team event refers to making up the team so that it can work as a whole in relation to the opposing team’s composition.


2020 ◽  
Vol 54 (10) ◽  
pp. 566-572 ◽  
Author(s):  
Sharief Hendricks ◽  
Kevin Till ◽  
Steve den Hollander ◽  
Trevor N Savage ◽  
Simon P Roberts ◽  
...  

Using an expert consensus-based approach, a rugby union Video Analysis Consensus (RUVAC) group was formed to develop a framework for video analysis research in rugby union. The aim of the framework is to improve the consistency of video analysis work in rugby union and help enhance the overall quality of future research in the sport. To reach consensus, a systematic review and Delphi method study design was used. After a systematic search of the literature, 17 articles were used to develop the final framework that described and defined key actions and events in rugby union (rugby). Thereafter, a group of researchers and practitioners with experience and expertise in rugby video analysis formed the RUVAC group. Each member of the group examined the framework of descriptors and definitions and rated their level of agreement on a 5-point agreement Likert scale (1: strongly disagree; 2: disagree; 3: neither agree or disagree; 4: agree; 5: strongly agree). The mean rating of agreement on the five-point scale (1: strongly disagree; 5: strongly agree) was 4.6 (4.3–4.9), 4.6 (4.4–4.9), 4.7 (4.5–4.9), 4.8 (4.6–5.0) and 4.8 (4.6–5.0) for the tackle, ruck, scrum, line-out and maul, respectively. The RUVAC group recommends using this consensus as the starting framework when conducting rugby video analysis research. Which variables to use (if not all) depends on the objectives of the study. Furthermore, the intention of this consensus is to help integrate video data with other data (eg, injury surveillance).


2020 ◽  
Vol 71 (7) ◽  
pp. 868-880
Author(s):  
Nguyen Hong Quan ◽  
Nguyen Thuy Binh ◽  
Tran Duc Long ◽  
Le Thi Lan

Along with the strong development of camera networks, a video analysis system has been become more and more popular and has been applied in various practical applications. In this paper, we focus on person re-identification (person ReID) task that is a crucial step of video analysis systems. The purpose of person ReID is to associate multiple images of a given person when moving in a non-overlapping camera network. Many efforts have been made to person ReID. However, most of studies on person ReID only deal with well-alignment bounding boxes which are detected manually and considered as the perfect inputs for person ReID. In fact, when building a fully automated person ReID system the quality of the two previous steps that are person detection and tracking may have a strong effect on the person ReID performance. The contribution of this paper are two-folds. First, a unified framework for person ReID based on deep learning models is proposed. In this framework, the coupling of a deep neural network for person detection and a deep-learning-based tracking method is used. Besides, features extracted from an improved ResNet architecture are proposed for person representation to achieve a higher ReID accuracy. Second, our self-built dataset is introduced and employed for evaluation of all three steps in the fully automated person ReID framework.


2014 ◽  
Vol 44 ◽  
pp. 161-170 ◽  
Author(s):  
Nicolas Courty ◽  
Pierre Allain ◽  
Clement Creusot ◽  
Thomas Corpetti

2018 ◽  
Vol 2 (2) ◽  
Author(s):  
Bernd Gössling ◽  
Desirée Daniel

The use of video analysis in Design-Based Research (DBR) seems to be promising, because the quality of video data matches the reality of educational fields. Educational fields are multidimensional and complex. And more than other types of data, video may capture, for example, the simultaneity of verbal and non-verbal interactions. This seems to be valuable in the quest for new insights and better designs of educational interventions. However, to date there has been limited use of video data in researching their design. This paper aims at reflecting how the benefits of video-based analysis may be utilised in DBR. Experiences with the collection and analysis of video data in a project to design self-organised learning (SOL) at a vocational school in Germany will be used as a case study to illustrate the type of findings that may feed into the DBR process. In this case, the project school had already introduced a sophisticated SOL model but was experiencing various implementation difficulties. Resolving issues like this requires insights into how exactly a concept is realised and what happens in the field. Therefore, video data on classroom interactions was gathered and sub-sequently analysed using the documentary method. This led to the reconstruction of two different types of orientation that were guiding the students when they dealt with their self-organised learning environment. In a subversive orientation, students playfully infiltrate the formal learning space with peer activities. In a confirming orientation, students stick to both, the (informal) rules of the (formal) learning arrangement and of the peer environment, thus expressing respect for the boundary between these two worlds. These findings have been used to redesign the SOL intervention.


Sign in / Sign up

Export Citation Format

Share Document