| 31. |
- Abdeljaber, Osama, et al.
(författare)
-
Extraction of Vehicle Turning Trajectories at Signalized Intersections Using Convolutional Neural Networks
- 2020
-
Ingår i: Arabian Journal for Science and Engineering. - Heidelberg : Springer. - 2193-567X .- 2191-4281 .- 1319-8025. ; 45, s. 8011-8025
-
Tidskriftsartikel (refereegranskat)abstract
- This paper aims at developing a convolutional neural network (CNN)-based tool that can automatically detect the left-turning vehicles (right-hand traffic rule) at signalized intersections and extract their trajectories from a recorded video. The proposed tool uses a region-based CNN trained over a limited number of video frames to detect moving vehicles. Kalman filters are then used to track the detected vehicles and extract their trajectories. The proposed tool achieved an acceptable accuracy level when verified against the manually extracted trajectories, with an average error of 16.5 cm. Furthermore, the trajectories extracted using the proposed vehicle tracking method were used to demonstrate the applicability of the minimum-jerk principle to reproduce variations in the vehicles’ paths. The effort presented in this paper can be regarded as a way forward toward maximizing the potential use of deep learning in traffic safety applications.
|
|
| 32. |
- Abebe, Solomon Akele, et al.
(författare)
-
Unfolding the Dynamics of Refugees’ Entrepreneurial Journey in the Aftermath of Forced Displacement
- 2023
-
Ingår i: The Palgrave Handbook of Global Migration in International Business. - Cham : Palgrave Macmillan. ; , s. 465-499
-
Bokkapitel (refereegranskat)abstract
- Despite their detrimental circumstances, the number of refugee-owned businesses is on the rise-a situation referred to as “the paradox of refugee entrepreneurship.” However, a key question is how refugees, having undergone extreme life disruption due to war, conflict, and forced displacement, fare as entrepreneurs in their host country. Extant studies do not address this question, as they primarily focus on factors determining refugees’ entrepreneurial entry. We extend current research, which predominantly focuses on the antecedents of refugee entrepreneurship, with a processual approach that captures refugees’ entrepreneurial journey. Drawing on inductive theory-building analysis and extensive data gathered from 40 in-depth interviews with 21 recently arrived Syrians in Sweden, our dynamic model provides a thorough examination of the tactics and procedures that refugees employ to create and grow their businesses. This research contributes to the advancement of existing theory by introducing a comprehensive model that delves into the intricacies of refugee entrepreneurship, outlining the distinct phases, detailing the underlying mechanisms for each phase, and identifying the critical factors that drive progress from one stage to the next. Our findings offer valuable insights for initiatives seeking to provide support and encouragement to refugee entrepreneurs, thereby fostering their success and empowering them to reach their full potential. © 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG
|
|
| 33. |
- Abella, Jaume, et al.
(författare)
-
SAFEXPLAIN : Safe and Explainable Critical Embedded Systems Based on AI
- 2023
-
Ingår i: DATE 23: Design, Automation And Test In Europe. - 9783981926378 ; , s. 1-6
-
Konferensbidrag (refereegranskat)abstract
- Deep Learning (DL) techniques are at the heart of most future advanced software functions in Critical Autonomous AI-based Systems (CAIS), where they also represent a major competitive factor. Hence, the economic success of CAIS industries (e.g., automotive, space, railway) depends on their ability to design, implement, qualify, and certify DL-based software products under bounded effort/cost. However, there is a fundamental gap between Functional Safety (FUSA) requirements on CAIS and the nature of DL solutions. This gap stems from the development process of DL libraries and affects high-level safety concepts such as (1) explainability and traceability, (2) suitability for varying safety requirements, (3) FUSA-compliant implementations, and (4) real-time constraints. As a matter of fact, the data-dependent and stochastic nature of DL algorithms clashes with current FUSA practice, which instead builds on deterministic, verifiable, and pass/fail test-based software. The SAFEXPLAIN project tackles these challenges and targets by providing a flexible approach to allow the certification - hence adoption - of DL-based solutions in CAIS building on: (1) DL solutions that provide end-to-end traceability, with specific approaches to explain whether predictions can be trusted and strategies to reach (and prove) correct operation, in accordance to certification standards; (2) alternative and increasingly sophisticated design safety patterns for DL with varying criticality and fault tolerance requirements; (3) DL library implementations that adhere to safety requirements; and (4) computing platform configurations, to regain determinism, and probabilistic timing analyses, to handle the remaining non-determinism. © 2023 EDAA.
|
|
| 34. |
- Abiri, Najmeh, et al.
(författare)
-
Establishing strong imputation performance of a denoising autoencoder in a wide range of missing data problems
- 2019
-
Ingår i: Neurocomputing. - Amsterdam : Elsevier BV. - 0925-2312 .- 1872-8286. ; 365, s. 137-146
-
Tidskriftsartikel (refereegranskat)abstract
- Dealing with missing data in data analysis is inevitable. Although powerful imputation methods that address this problem exist, there is still much room for improvement. In this study, we examined single imputation based on deep autoencoders, motivated by the apparent success of deep learning to efficiently extract useful dataset features. We have developed a consistent framework for both training and imputation. Moreover, we benchmarked the results against state-of-the-art imputation methods on different data sizes and characteristics. The work was not limited to the one-type variable dataset; we also imputed missing data with multi-type variables, e.g., a combination of binary, categorical, and continuous attributes. To evaluate the imputation methods, we randomly corrupted the complete data, with varying degrees of corruption, and then compared the imputed and original values. In all experiments, the developed autoencoder obtained the smallest error for all ranges of initial data corruption.
|
|
| 35. |
- Abiri, Najmeh, et al.
(författare)
-
Variational auto-encoders with Student’s t-prior
- 2019
-
Ingår i: ESANN 2019 - Proceedings : The 27th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning - The 27th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning. - Bruges : ESANN. - 9782875870650
-
Konferensbidrag (refereegranskat)abstract
- We propose a new structure for the variational auto-encoders (VAEs) prior, with the weakly informative multivariate Student’s t-distribution. In the proposed model all distribution parameters are trained, thereby allowing for a more robust approximation of the underlying data distribution. We used Fashion-MNIST data in two experiments to compare the proposed VAEs with the standard Gaussian priors. Both experiments showed a better reconstruction of the images with VAEs using Student’s t-prior distribution.
|
|
| 36. |
- Aboelwafa, Mariam M. N., et al.
(författare)
-
A Machine-Learning-Based Technique for False Data Injection Attacks Detection in Industrial IoT
- 2020
-
Ingår i: IEEE Internet of Things Journal. - Piscataway : Institute of Electrical and Electronics Engineers (IEEE). - 2327-4662. ; 7:9, s. 8462-8471
-
Tidskriftsartikel (refereegranskat)abstract
- The accelerated move toward the adoption of the Industrial Internet-of-Things (IIoT) paradigm has resulted in numerous shortcomings as far as security is concerned. One of the IIoT affecting critical security threats is what is termed as the false data injection (FDI) attack. The FDI attacks aim to mislead the industrial platforms by falsifying their sensor measurements. FDI attacks have successfully overcome the classical threat detection approaches. In this article, we present a novel method of FDI attack detection using autoencoders (AEs). We exploit the sensor data correlation in time and space, which in turn can help identify the falsified data. Moreover, the falsified data are cleaned using the denoising AEs (DAEs). Performance evaluation proves the success of our technique in detecting FDI attacks. It also significantly outperforms a support vector machine (SVM)-based approach used for the same purpose. The DAE data cleaning algorithm is also shown to be very effective in recovering clean data from corrupted (attacked) data. © 2014 IEEE.
|
|
| 37. |
- Abrahamsson, Cristian, et al.
(författare)
-
Content, interest and the role of engagement : experienced science teachers discuss
- 2023
-
Ingår i: Physics Education. - Bristol : Institute of Physics Publishing (IOPP). - 0031-9120 .- 1361-6552. ; 58:6
-
Tidskriftsartikel (refereegranskat)abstract
- How do science teachers perceive student engagement and its importance for teaching and what strategies do they use to create it? When 21 experienced science teachers in 4 focus groups discussed these questions, they brought up behavioural aspects, but also less visible emotional and cognitive aspects, as well as reciprocal aspects of teacher and student engagement. One teacher described engagement as 'the oil in the machinery' during lessons. Which role does the curricular content play? Well aware that some topics are seen as more directly interesting by students, teachers connect to these, but also use hooks, including lively demonstrations, role play and connections to the outside world. In this way, they aim to generate situational interest and engagement also in topics that are often viewed as less interesting, including atoms and molecules. These experienced teachers describe how they adapt their teaching to the group also in real time, based on the degree of engagement exhibited by the students. © 2023 The Author(s). Published by IOP Publishing Ltd.
|
|
| 38. |
- Abrahamsson, Cristian, et al.
(författare)
-
En Delfistudie om lärares uppfattning av elevengagemang i NO-undervisningen
- 2019
-
Ingår i: NorDiNa. - Oslo : Naturfagsenteret / Norwegian Centre for Science Education. - 1504-4556 .- 1894-1257. ; 15:2, s. 128-144
-
Tidskriftsartikel (refereegranskat)abstract
- What happens in a science classroom where students are engaged and how do teachers observe and interpret student engagement? This article highlights teachers’ perspective on students’ engagement in science education and to what extent it is connected to the scientific content. This approach complements earlier research which focuses mostly on students’ attitude towards science education and their interest in various topics in science.The findings are based on a three-stage Delphi survey distributed to 39 expert science teachers. The results shows science education with a range of different perspectives and that most teachers do not perceive any direct connection between specific science topics and the students’ engagement. The survey also shows that teachers to a high level interpret students’ emotional expressions and academic behavior as engagement rather than their cognitive behavior.
|
|
| 39. |
- Abrahamsson, Kajsa H., 1956, et al.
(författare)
-
Patients´views on periodontal disease; attitutes to oral health and expectancy of periodontal treatment: a qualitative interview study
- 2008
-
Ingår i: Oral Health & Preventive Dentistry. - New Malden, Surry : Quintessence Publishing Co. Ltd.. - 1602-1622 .- 1757-9996. ; 6:3, s. 209-216
-
Tidskriftsartikel (refereegranskat)abstract
- Purpose: The aim of the study was to explore and gain an understanding of patients' views on their periodontal conditions, their perceived impact of periodontitis on daily life, as well as their attitudes to oral health and expectations of treatment. Materials and Methods: The study subjects were patients with chronic periodontitis, who had been referred to a specialist clinic. The constant comparative method for grounded theory was used to collect and analyse the data. Audiotaped, open-ended interviews were conducted after periodontal examination, but before treatment. The interviews were transcribed verbatim and consecutively analysed in hierarchical coding processes and continued until saturation was reached (n = 17). In the analysis, a conceptual model that outlined the steps involved in the diagnosis of periodontitis was generated. The core concept of the model, keeping up appearance and self-esteem, was related to the following four additional categories and their dimensions; doing what you have to do - trying to live up to the norm, suddenly having a shameful and disabling disease, feeling deserted and in the hands of an authority, and investing all in a treatment with an unpredictable outcome. Results: The results illustrated that subjects diagnosed with chronic periodontitis felt ashamed and were willing to invest all they had in terms of time, effort and financial resources to become healthy and to maintain their self-esteem. However, they perceived a low degree of control over treatment decisions and treatment outcome. Conclusions: The results demonstrate the vulnerability of patients diagnosed with chronic periodontitis and emphasise the importance of communication in dentistry.
|
|
| 40. |
|
|
