You searched for subject:(Attention Mechanism).
Showing records 1 – 17 of
17 total matches.
No search limiters apply to these results.
NSYSU
1.
Hsiao, Po-Wei.
Deep Neural Networks and Ensemble Learning with Application to Speech Emotion Recognition.
Degree: Master, Computer Science and Engineering, 2018, NSYSU
URL: http://etd.lib.nsysu.edu.tw/ETD-db/ETD-search/view_etd?URN=etd-0730118-100155 
► This study uses deep neural networks to construct the static and dynamic speech emotion recognition systems and integrates the static and dynamic models by ensemble…
(more)
▼ This study uses deep neural networks to construct the static and dynamic speech emotion recognition systems and integrates the static and dynamic models by ensemble learning. The static model is based on multi-layer perceptron (MLP) and convolutional neural network (CNN). The dynamic model is based on recurrent neural network (RNN). Our CNN recognizer learns to focus on salient parts of signal by the
attention mechanism, and promotes competition among a set of multi-scale convolutional filters by multi-scale convolution module. The RNN recognizer also incorporates the
attention mechanism to learn to focus on the informative segments. We adopt a skew-robust training criterion to deal with unbalanced data. We also exploit a two-pass teacher-student training scheme to deal with the issue of noisy labels. The proposed speech emotion recognition systems are evaluated on the FAU-Aibo corpus, using the tasks as defined in the Interspeech 2009 Emotion Challenge classifier sub-challenge, with the performance measure of unweighted average recall rate (UA). Our MLP and CNN models achieve 46.2% and 46.4% UA respectively, and our dynamic model with deep RNN achieves 47.2% UA, surpassing the previous best mark of 46.4%. Further, an ensemble learning implemented by interpolation that combines the static and dynamic models achieves 50.5% UA, breaking the 50.0% barrier on the FAU-Aibo tasks for the first time since the Challenge is posted.
Advisors/Committee Members: Shen-Fu Hsiao (chair), Chung-Nan Lee (chair), Chia-Ping Chen (committee member), Yun-Nan Chang (chair), Chung-Hsien Wu (chair).
Subjects/Keywords: attention mechanism; speech emotion recognition; ensemble learning; deep neural networks
Record Details
Similar Records
Cite
Share »
Record Details
Similar Records
Cite
« Share
❌
APA ·
Chicago ·
MLA ·
Vancouver ·
CSE |
Export
to Zotero / EndNote / Reference
Manager
APA (6th Edition):
Hsiao, P. (2018). Deep Neural Networks and Ensemble Learning with Application to Speech Emotion Recognition. (Thesis). NSYSU. Retrieved from http://etd.lib.nsysu.edu.tw/ETD-db/ETD-search/view_etd?URN=etd-0730118-100155
Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation
Chicago Manual of Style (16th Edition):
Hsiao, Po-Wei. “Deep Neural Networks and Ensemble Learning with Application to Speech Emotion Recognition.” 2018. Thesis, NSYSU. Accessed March 01, 2021.
http://etd.lib.nsysu.edu.tw/ETD-db/ETD-search/view_etd?URN=etd-0730118-100155.
Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation
MLA Handbook (7th Edition):
Hsiao, Po-Wei. “Deep Neural Networks and Ensemble Learning with Application to Speech Emotion Recognition.” 2018. Web. 01 Mar 2021.
Vancouver:
Hsiao P. Deep Neural Networks and Ensemble Learning with Application to Speech Emotion Recognition. [Internet] [Thesis]. NSYSU; 2018. [cited 2021 Mar 01].
Available from: http://etd.lib.nsysu.edu.tw/ETD-db/ETD-search/view_etd?URN=etd-0730118-100155.
Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation
Council of Science Editors:
Hsiao P. Deep Neural Networks and Ensemble Learning with Application to Speech Emotion Recognition. [Thesis]. NSYSU; 2018. Available from: http://etd.lib.nsysu.edu.tw/ETD-db/ETD-search/view_etd?URN=etd-0730118-100155
Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation
Delft University of Technology
2.
Li, Jiahui (author).
Attention-Aware Age-Agnostic Visual Place Recognition.
Degree: 2019, Delft University of Technology
URL: http://resolver.tudelft.nl/uuid:250d37a9-bc0d-4f8f-8d1a-d31a98dc22d7
► A cross-domain visual place recognition (VPR) task is proposed in this work, i.e., matching images of the same architectures depicted in different domains. VPR is…
(more)
▼ A cross-domain visual place recognition (VPR) task is proposed in this work, i.e., matching images of the same architectures depicted in different domains. VPR is commonly treated as an image retrieval task, where a query image from an unknown location is matched with relevant instances from geo-tagged gallery database. Different from conventional VPR settings where the query images and gallery images come from the same domain, we propose a more common but challenging setup where the query images are collected under a new unseen condition. The two domains involved in this work are contemporary street view images of Amsterdam from the Mapillary dataset (source domain) and historical images of the same city from Beeldbank dataset (target domain). We tailored an age-invariant feature learning CNN that can focus on domain invariant objects and learn to match images based on a weakly supervised ranking loss. We propose an
attention aggregation module that is robust to domain discrepancy between the train and the test data. Further, a multi-kernel maximum mean discrepancy (MK-MMD) domain adaptation loss is adopted to improve the cross-domain ranking performance. Both
attention and adaptation modules are unsupervised while the ranking loss uses weak supervision. Visual inspection shows that the
attention module focuses on built forms while the dramatically changing environment are less weighed. Our proposed CNN achieves state of the art results (99% accuracy) on the single-domain VPR task and 20% accuracy at its best on the cross-domain VPR task, revealing the difficulty of age-invariant VPR.
Advisors/Committee Members: van Gemert, Jan (mentor), Khademi, Seyran (mentor), Wang, Ziqi (mentor), Reinders, Marcel (graduation committee), Nan, Liangliang (graduation committee), Delft University of Technology (degree granting institution).
Subjects/Keywords: Computer Vision; Domain Adaptation; Image Matching; Attention Mechanism
Record Details
Similar Records
Cite
Share »
Record Details
Similar Records
Cite
« Share
❌
APA ·
Chicago ·
MLA ·
Vancouver ·
CSE |
Export
to Zotero / EndNote / Reference
Manager
APA (6th Edition):
Li, J. (. (2019). Attention-Aware Age-Agnostic Visual Place Recognition. (Masters Thesis). Delft University of Technology. Retrieved from http://resolver.tudelft.nl/uuid:250d37a9-bc0d-4f8f-8d1a-d31a98dc22d7
Chicago Manual of Style (16th Edition):
Li, Jiahui (author). “Attention-Aware Age-Agnostic Visual Place Recognition.” 2019. Masters Thesis, Delft University of Technology. Accessed March 01, 2021.
http://resolver.tudelft.nl/uuid:250d37a9-bc0d-4f8f-8d1a-d31a98dc22d7.
MLA Handbook (7th Edition):
Li, Jiahui (author). “Attention-Aware Age-Agnostic Visual Place Recognition.” 2019. Web. 01 Mar 2021.
Vancouver:
Li J(. Attention-Aware Age-Agnostic Visual Place Recognition. [Internet] [Masters thesis]. Delft University of Technology; 2019. [cited 2021 Mar 01].
Available from: http://resolver.tudelft.nl/uuid:250d37a9-bc0d-4f8f-8d1a-d31a98dc22d7.
Council of Science Editors:
Li J(. Attention-Aware Age-Agnostic Visual Place Recognition. [Masters Thesis]. Delft University of Technology; 2019. Available from: http://resolver.tudelft.nl/uuid:250d37a9-bc0d-4f8f-8d1a-d31a98dc22d7
University of Melbourne
3.
Chen, Jian.
The role of attention in subitizing.
Degree: 2019, University of Melbourne
URL: http://hdl.handle.net/11343/230741
► This thesis aimed to address a long-standing question: “Why are small numbers of items enumerated differently from large numbers of items?” (Trick & Pylyshyn, 1993,…
(more)
▼ This thesis aimed to address a long-standing question: “Why are small numbers of items enumerated differently from large numbers of items?” (Trick & Pylyshyn, 1993, p. 331). Specifically, this thesis examined whether the processing of small sets requires attention processes and, if so, how attention plays a role in subitizing. Answers to this question are central to debates about the nature of contemporary models of numerical cognition.
In this thesis, attention was manipulated in an enumeration task using the Posner Cueing paradigm. The physical properties of to-be-enumerated objects in the experiments were well-controlled. Different display times of objects were used in the experiments. An internal mini meta-analysis was run to synthesize the attention effects in these experiments. A meta-analysis was also conducted to evaluate the observed attention effects in previous studies. Moreover, to investigate how attention plays a role in the subitizing range, experiments was conducted to compare the attention effect between an enumeration task and a control task.
Findings from the experiments reported in this thesis suggested a robust attention effect in the subitizing range: there was no dichotomy in attention between subitizing and counting. These findings imply that subitizing processing requires attention. The meta- analysis evaluating previous studies also suggested a robust attention effect in the subitizing range. Moreover, the findings reported in this thesis suggested that attention does not specifically affect numerical processing. Instead, attention played an important role in general cognitive processes in the subitizing range. A tentative model was proposed to explain the mechanisms underlying visual enumeration, especially in the subitizing range.
In summary, these findings suggest that enumeration in the subitizing range requires attention. Specifically, attention is required for general processes in the subitizing range.
Subjects/Keywords: subitizing; counting; enumeration; attention; Posner cueing paradigm; visual enumeration mechanism
Record Details
Similar Records
Cite
Share »
Record Details
Similar Records
Cite
« Share
❌
APA ·
Chicago ·
MLA ·
Vancouver ·
CSE |
Export
to Zotero / EndNote / Reference
Manager
APA (6th Edition):
Chen, J. (2019). The role of attention in subitizing. (Doctoral Dissertation). University of Melbourne. Retrieved from http://hdl.handle.net/11343/230741
Chicago Manual of Style (16th Edition):
Chen, Jian. “The role of attention in subitizing.” 2019. Doctoral Dissertation, University of Melbourne. Accessed March 01, 2021.
http://hdl.handle.net/11343/230741.
MLA Handbook (7th Edition):
Chen, Jian. “The role of attention in subitizing.” 2019. Web. 01 Mar 2021.
Vancouver:
Chen J. The role of attention in subitizing. [Internet] [Doctoral dissertation]. University of Melbourne; 2019. [cited 2021 Mar 01].
Available from: http://hdl.handle.net/11343/230741.
Council of Science Editors:
Chen J. The role of attention in subitizing. [Doctoral Dissertation]. University of Melbourne; 2019. Available from: http://hdl.handle.net/11343/230741
Brno University of Technology
4.
Hradil, Jaromír.
Rozpoznávání řeči překrývajících se řečníků pomocí neuronových sítí: Recognition of Multi-Talker Overlapping Speech Using Neural Networks.
Degree: 2020, Brno University of Technology
URL: http://hdl.handle.net/11012/191516
► This work deals with the speech recognition of overlapping speakers using a neural network. It examines the problem of speech recognition from multiple speakers and…
(more)
▼ This work deals with the speech recognition of overlapping speakers using a neural network. It examines the problem of speech recognition from multiple speakers and the ways in which this problem is solved. Specifically, in addition to traditional components such as convolutional neural networks, LSTM, etc., it is also an application of special components:
attention mechanism and gated convolution. And also the application of a technique called permutation invariant training. Part of this work is to apply these approaches to assigned training data, which consists of artificially created mixtures of two speakers reading articles from the Wall Street Journal. The next step was to train the respective architectures using the combinations of the elements mentioned above. The models in this work replace the acoustic model. There were two architectures using different types of
attention mechanism and one without it. Experiments have shown that architectures using the
attention mechanism in this type of task have not surpassed more traditional architecture by suffering from gated convolution. Nevertheless, they showed potential.
Advisors/Committee Members: Žmolíková, Kateřina (advisor), Švec, Ján (referee).
Subjects/Keywords: rozpoznávání řeči; neuronové sítě; attention mechanismus; překrývající se řeč; speech recognition; neural networks; attention mechanism; overlapping speech
Record Details
Similar Records
Cite
Share »
Record Details
Similar Records
Cite
« Share
❌
APA ·
Chicago ·
MLA ·
Vancouver ·
CSE |
Export
to Zotero / EndNote / Reference
Manager
APA (6th Edition):
Hradil, J. (2020). Rozpoznávání řeči překrývajících se řečníků pomocí neuronových sítí: Recognition of Multi-Talker Overlapping Speech Using Neural Networks. (Thesis). Brno University of Technology. Retrieved from http://hdl.handle.net/11012/191516
Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation
Chicago Manual of Style (16th Edition):
Hradil, Jaromír. “Rozpoznávání řeči překrývajících se řečníků pomocí neuronových sítí: Recognition of Multi-Talker Overlapping Speech Using Neural Networks.” 2020. Thesis, Brno University of Technology. Accessed March 01, 2021.
http://hdl.handle.net/11012/191516.
Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation
MLA Handbook (7th Edition):
Hradil, Jaromír. “Rozpoznávání řeči překrývajících se řečníků pomocí neuronových sítí: Recognition of Multi-Talker Overlapping Speech Using Neural Networks.” 2020. Web. 01 Mar 2021.
Vancouver:
Hradil J. Rozpoznávání řeči překrývajících se řečníků pomocí neuronových sítí: Recognition of Multi-Talker Overlapping Speech Using Neural Networks. [Internet] [Thesis]. Brno University of Technology; 2020. [cited 2021 Mar 01].
Available from: http://hdl.handle.net/11012/191516.
Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation
Council of Science Editors:
Hradil J. Rozpoznávání řeči překrývajících se řečníků pomocí neuronových sítí: Recognition of Multi-Talker Overlapping Speech Using Neural Networks. [Thesis]. Brno University of Technology; 2020. Available from: http://hdl.handle.net/11012/191516
Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation
Lehigh University
5.
Nguyen, Chanh.
Two approaches to defend against adversarial examples: Attention-based and Certificate-based.
Degree: MS, Computer Science, 2018, Lehigh University
URL: https://preserve.lehigh.edu/etd/4363
► In this paper, we present two different novel approaches to defend against adversarial examples in neural networks: attention-based against pixel-based attack and certificate-based against spatially…
(more)
▼ In this paper, we present two different novel approaches to defend against adversarial examples in neural networks:
attention-based against pixel-based attack and certificate-based against spatially transformed attack. We discuss the vulnerability of neural networks for adversarial examples, which significantly hinders their application in security-critical domains. We detail several popular pixel-based methods of attacking a model. We then walk through current defense methods and note that they can often be circumvented by adaptive adversaries. For the first contribution, we take a completely different route by leveraging the definition of adversarial inputs: while deceiving for deep neural networks, they are barely discernible for human visions. Building upon recent advances in interpretable models, we construct a new detection framework that contrasts an input’s interpretation against its classification. We validate the efficacy of this framework through extensive experiments using benchmark datasets and attacks. We believe that this work opens a new direction for designing adversarial input detection methods. As for the second contribution, we discuss a completely different approach to generate adversarial examples, based on the spatial transformation of an input image. We then extend a currently proposed certificate framework to this setting and show that the certificate can improve the resilience of a network against adversarial spatial transformation.
Advisors/Committee Members: Wang, Ting.
Subjects/Keywords: Adversarial examples; Attention mechanism; Deep learning; Machine learning; Provable security; Computer Sciences
Record Details
Similar Records
Cite
Share »
Record Details
Similar Records
Cite
« Share
❌
APA ·
Chicago ·
MLA ·
Vancouver ·
CSE |
Export
to Zotero / EndNote / Reference
Manager
APA (6th Edition):
Nguyen, C. (2018). Two approaches to defend against adversarial examples: Attention-based and Certificate-based. (Thesis). Lehigh University. Retrieved from https://preserve.lehigh.edu/etd/4363
Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation
Chicago Manual of Style (16th Edition):
Nguyen, Chanh. “Two approaches to defend against adversarial examples: Attention-based and Certificate-based.” 2018. Thesis, Lehigh University. Accessed March 01, 2021.
https://preserve.lehigh.edu/etd/4363.
Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation
MLA Handbook (7th Edition):
Nguyen, Chanh. “Two approaches to defend against adversarial examples: Attention-based and Certificate-based.” 2018. Web. 01 Mar 2021.
Vancouver:
Nguyen C. Two approaches to defend against adversarial examples: Attention-based and Certificate-based. [Internet] [Thesis]. Lehigh University; 2018. [cited 2021 Mar 01].
Available from: https://preserve.lehigh.edu/etd/4363.
Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation
Council of Science Editors:
Nguyen C. Two approaches to defend against adversarial examples: Attention-based and Certificate-based. [Thesis]. Lehigh University; 2018. Available from: https://preserve.lehigh.edu/etd/4363
Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation
York University
6.
Abbaszadegan, Mahdieh.
An Encoder-Decoder Based Basecaller for Nanopore DNA Sequencing.
Degree: MASc - Master of Applied Science, Electrical and Computer Engineering, 2019, York University
URL: http://hdl.handle.net/10315/36268
► Nanopore DNA sequencing is a method in which DNA bases are determined (basecalled) using electric current signals generated by passing DNA through nanopore sensors. The…
(more)
▼ Nanopore DNA sequencing is a method in which DNA bases are determined (basecalled) using electric current signals generated by passing DNA through nanopore sensors. The raw measured signals can be aggregated into event data presenting new bases entering the nanopore. This thesis has two contributions. First, we implemented RNN-based single- and double-strand basecallers for simulated event data to analyze the effect of signal noise. As the SNR decreased from 20 dB to 5 dB, the accuracy of the single-strand basecaller dropped 9% while the accuracy of double-strand basecaller only dropped 0.5%. Second, we implemented an end-to-end single-strand basecaller, directly processing the raw signal using an encoder-decoder model with
attention instead of the CTC-style approach used in available basecallers. We achieved an accuracy of 81.9% for a viral sample and an accuracy of 90.9% for a bacterial sample. Our accuracy is comparable to state-of-the-art basecallers with a considerably smaller model.
Advisors/Committee Members: Magierowski, Sebastian (advisor).
Subjects/Keywords: Computer science; DNA Sequencing; Nanopore Sequencing; Deep Learning; Recurrent Neural Networks; Seq2seq; Attention Mechanism
Record Details
Similar Records
Cite
Share »
Record Details
Similar Records
Cite
« Share
❌
APA ·
Chicago ·
MLA ·
Vancouver ·
CSE |
Export
to Zotero / EndNote / Reference
Manager
APA (6th Edition):
Abbaszadegan, M. (2019). An Encoder-Decoder Based Basecaller for Nanopore DNA Sequencing. (Masters Thesis). York University. Retrieved from http://hdl.handle.net/10315/36268
Chicago Manual of Style (16th Edition):
Abbaszadegan, Mahdieh. “An Encoder-Decoder Based Basecaller for Nanopore DNA Sequencing.” 2019. Masters Thesis, York University. Accessed March 01, 2021.
http://hdl.handle.net/10315/36268.
MLA Handbook (7th Edition):
Abbaszadegan, Mahdieh. “An Encoder-Decoder Based Basecaller for Nanopore DNA Sequencing.” 2019. Web. 01 Mar 2021.
Vancouver:
Abbaszadegan M. An Encoder-Decoder Based Basecaller for Nanopore DNA Sequencing. [Internet] [Masters thesis]. York University; 2019. [cited 2021 Mar 01].
Available from: http://hdl.handle.net/10315/36268.
Council of Science Editors:
Abbaszadegan M. An Encoder-Decoder Based Basecaller for Nanopore DNA Sequencing. [Masters Thesis]. York University; 2019. Available from: http://hdl.handle.net/10315/36268
Brno University of Technology
7.
Dronzeková, Michaela.
Analýza polygonálních modelů pomocí neuronových sítí: Analysis of Polygonal Models Using Neural Networks.
Degree: 2020, Brno University of Technology
URL: http://hdl.handle.net/11012/192482
► This thesis deals with rotation estimation of 3D model of human jaw. It describes and compares methods for direct analysis od 3D models as well…
(more)
▼ This thesis deals with rotation estimation of 3D model of human jaw. It describes and compares methods for direct analysis od 3D models as well as method to analyze model using rasterization. To evaluate perfomance of proposed method, a metric that computes number of cases when prediction was less than 30° from ground truth is used. Proposed method that uses rasterization, takes three x-ray views of model as an input and processes it with convolutional network. It achieves best preformance, 99% with described metric. Method to directly analyze polygonal model as a sequence uses
attention mechanism to do so and was inspired by transformer architecture. A special pooling function was proposed for this network that decreases memory requirements of the network. This method achieves 88%, but does not use rasterization and can process polygonal model directly. It is not as good as rasterization method with x-ray display, byt it is better than rasterization method with model not rendered as x-ray. The last method uses graph representation of mesh. Graph network had problems with overfitting, that is why it did not get good results and I think this method is not very suitable for analyzing plygonal model.
Advisors/Committee Members: Kodym, Oldřich (advisor), Zemčík, Pavel (referee).
Subjects/Keywords: polygoniálne modely; neurónové siete; odhad rotácie; attention mechanizmus; transformer; grafové konvolučné siete; polygonal models; neural networks; rotation estimation; attention mechanism; transformer; graph convolutional network
Record Details
Similar Records
Cite
Share »
Record Details
Similar Records
Cite
« Share
❌
APA ·
Chicago ·
MLA ·
Vancouver ·
CSE |
Export
to Zotero / EndNote / Reference
Manager
APA (6th Edition):
Dronzeková, M. (2020). Analýza polygonálních modelů pomocí neuronových sítí: Analysis of Polygonal Models Using Neural Networks. (Thesis). Brno University of Technology. Retrieved from http://hdl.handle.net/11012/192482
Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation
Chicago Manual of Style (16th Edition):
Dronzeková, Michaela. “Analýza polygonálních modelů pomocí neuronových sítí: Analysis of Polygonal Models Using Neural Networks.” 2020. Thesis, Brno University of Technology. Accessed March 01, 2021.
http://hdl.handle.net/11012/192482.
Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation
MLA Handbook (7th Edition):
Dronzeková, Michaela. “Analýza polygonálních modelů pomocí neuronových sítí: Analysis of Polygonal Models Using Neural Networks.” 2020. Web. 01 Mar 2021.
Vancouver:
Dronzeková M. Analýza polygonálních modelů pomocí neuronových sítí: Analysis of Polygonal Models Using Neural Networks. [Internet] [Thesis]. Brno University of Technology; 2020. [cited 2021 Mar 01].
Available from: http://hdl.handle.net/11012/192482.
Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation
Council of Science Editors:
Dronzeková M. Analýza polygonálních modelů pomocí neuronových sítí: Analysis of Polygonal Models Using Neural Networks. [Thesis]. Brno University of Technology; 2020. Available from: http://hdl.handle.net/11012/192482
Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation
Leiden University
8.
Lottes, Niklas R.L.
Does sustained attention mediate the relationship between mindfulness and depression and anxiety symptoms?.
Degree: 2015, Leiden University
URL: http://hdl.handle.net/1887/35179
Subjects/Keywords: Mindfulness; Sustained attention; Psychological distress; Working mechanism
Record Details
Similar Records
Cite
Share »
Record Details
Similar Records
Cite
« Share
❌
APA ·
Chicago ·
MLA ·
Vancouver ·
CSE |
Export
to Zotero / EndNote / Reference
Manager
APA (6th Edition):
Lottes, N. R. L. (2015). Does sustained attention mediate the relationship between mindfulness and depression and anxiety symptoms?. (Masters Thesis). Leiden University. Retrieved from http://hdl.handle.net/1887/35179
Chicago Manual of Style (16th Edition):
Lottes, Niklas R L. “Does sustained attention mediate the relationship between mindfulness and depression and anxiety symptoms?.” 2015. Masters Thesis, Leiden University. Accessed March 01, 2021.
http://hdl.handle.net/1887/35179.
MLA Handbook (7th Edition):
Lottes, Niklas R L. “Does sustained attention mediate the relationship between mindfulness and depression and anxiety symptoms?.” 2015. Web. 01 Mar 2021.
Vancouver:
Lottes NRL. Does sustained attention mediate the relationship between mindfulness and depression and anxiety symptoms?. [Internet] [Masters thesis]. Leiden University; 2015. [cited 2021 Mar 01].
Available from: http://hdl.handle.net/1887/35179.
Council of Science Editors:
Lottes NRL. Does sustained attention mediate the relationship between mindfulness and depression and anxiety symptoms?. [Masters Thesis]. Leiden University; 2015. Available from: http://hdl.handle.net/1887/35179
University of Tennessee – Knoxville
9.
Rahimpour, Alireza.
Attention Mechanism for Recognition in Computer Vision.
Degree: 2019, University of Tennessee – Knoxville
URL: https://trace.tennessee.edu/utk_graddiss/5592
► It has been proven that humans do not focus their attention on an entire scene at once when they perform a recognition task. Instead, they…
(more)
▼ It has been proven that humans do not focus their attention on an entire scene at once when they perform a recognition task. Instead, they pay attention to the most important parts of the scene to extract the most discriminative information. Inspired by this observation, in this dissertation, the importance of attention mechanism in recognition tasks in computer vision is studied by designing novel attention-based models. In specific, four scenarios are investigated that represent the most important aspects of attention mechanism.First, an attention-based model is designed to reduce the visual features' dimensionality by selectively processing only a small subset of the data. We study this aspect of the attention mechanism in a framework based on object recognition in distributed camera networks. Second, an attention-based image retrieval system (i.e., person re-identification) is proposed which learns to focus on the most discriminative regions of the person's image and process those regions with higher computation power using a deep convolutional neural network. Furthermore, we show how visualizing the attention maps can make deep neural networks more interpretable. In other words, by visualizing the attention maps we can observe the regions of the input image where the neural network relies on, in order to make a decision. Third, a model for estimating the importance of the objects in a scene based on a given task is proposed. More specifically, the proposed model estimates the importance of the road users that a driver (or an autonomous vehicle) should pay attention to in a driving scenario in order to have safe navigation. In this scenario, the attention estimation is the final output of the model. Fourth, an attention-based module and a new loss function in a meta-learning based few-shot learning system is proposed in order to incorporate the context of the task into the feature representations of the samples and increasing the few-shot recognition accuracy.In this dissertation, we showed that attention can be multi-facet and studied the attention mechanism from the perspectives of feature selection, reducing the computational cost, interpretable deep learning models, task-driven importance estimation, and context incorporation. Through the study of four scenarios, we further advanced the field of where ''attention is all you need''.
Subjects/Keywords: Attention mechanism; Machine learning; Deep Learning; Computer Vision; Meta-learning; Image retrieval and matching; Person Re-identification; Object detection; Feature selection.
Record Details
Similar Records
Cite
Share »
Record Details
Similar Records
Cite
« Share
❌
APA ·
Chicago ·
MLA ·
Vancouver ·
CSE |
Export
to Zotero / EndNote / Reference
Manager
APA (6th Edition):
Rahimpour, A. (2019). Attention Mechanism for Recognition in Computer Vision. (Doctoral Dissertation). University of Tennessee – Knoxville. Retrieved from https://trace.tennessee.edu/utk_graddiss/5592
Chicago Manual of Style (16th Edition):
Rahimpour, Alireza. “Attention Mechanism for Recognition in Computer Vision.” 2019. Doctoral Dissertation, University of Tennessee – Knoxville. Accessed March 01, 2021.
https://trace.tennessee.edu/utk_graddiss/5592.
MLA Handbook (7th Edition):
Rahimpour, Alireza. “Attention Mechanism for Recognition in Computer Vision.” 2019. Web. 01 Mar 2021.
Vancouver:
Rahimpour A. Attention Mechanism for Recognition in Computer Vision. [Internet] [Doctoral dissertation]. University of Tennessee – Knoxville; 2019. [cited 2021 Mar 01].
Available from: https://trace.tennessee.edu/utk_graddiss/5592.
Council of Science Editors:
Rahimpour A. Attention Mechanism for Recognition in Computer Vision. [Doctoral Dissertation]. University of Tennessee – Knoxville; 2019. Available from: https://trace.tennessee.edu/utk_graddiss/5592
KTH
10.
Navarro, Abgeiba Yaroslava Isunza.
Evaluation of Attention Mechanisms for Just-In-Time Software Defect Prediction.
Degree: Electrical Engineering and Computer Science (EECS), 2020, KTH
URL: http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-288724
► Just-In-Time Software Defect Prediction (JIT-DP) focuses on predicting errors in software at change-level with the objective of helping developers identify defects while the development…
(more)
▼ Just-In-Time Software Defect Prediction (JIT-DP) focuses on predicting errors in software at change-level with the objective of helping developers identify defects while the development process is still ongoing, and improving the quality of software applications. This work studies deep learning techniques by applying attention mechanisms that have been successful in, among others, Natural Language Processing (NLP) tasks. We introduce two networks named Convolutional Neural Network with Bidirectional Attention (BACNN) and Bidirectional Attention Code Network (BACoN) that employ a bi-directional attention mechanism between the code and message of a software change. Furthermore, we examine BERT [17] and RoBERTa [57] attention architectures for JIT-DP. More specifically, we study the effectiveness of the aforementioned attention-based models to predict defective commits compared to the current state of the art, DeepJIT [37] and TLEL [101]. Our experiments evaluate the models by using software changes from the OpenStack open source project. The results showed that attention-based networks outperformed the baseline models in terms of accuracy in the different evaluation settings. The attention-based models, particularly BERT and RoBERTa architectures, demonstrated promising results in identifying defective software changes and proved to be effective in predicting defects in changes of new software releases.
Just-In-Time Defect Prediction (JIT-DP) fokuserar på att förutspå fel i mjukvara vid ändringar i koden, med målet att hjälpa utvecklare att identifiera defekter medan utvecklingsprocessen fortfarande är pågående, och att förbättra kvaliteten hos applikationsprogramvara. Detta arbete studerar djupinlärningstekniker genom att tillämpa attentionmekanismer som har varit framgångsrika inom, bland annat, språkteknologi (NLP). Vi introducerar två nätverk vid namn Convolutional Neural Network with Bidirectional Attention (BACNN), och Bidirectional Attention Code Network (BACoN), som använder en tvåriktad attentionmekanism mellan koden och meddelandet om en mjukvaruändring. Dessutom undersöker vi BERT [17] och RoBERTa [57], attentionarkitekturer för JIT-DP. Mer specifikt studerar vi hur effektivt dessa attentionbaserade modeller kan förutspå defekta ändringar, och jämför dem med de bästa tillgängliga arkitekturerna DeePJIT [37] och TLEL [101]. Våra experiment utvärderar modellerna genom att använda mjukvaruändringar från det öppna källkodsprojektet OpenStack. Våra resultat visar att attentionbaserade nätverk överträffar referensmodellen sett till träffsäkerheten i de olika scenarierna. De attentionbaserade modellerna, framför allt BERT och RoBERTa, demonstrerade lovade resultat när det kommer till att identifiera defekta mjukvaruändringar och visade sig vara effektiva på att förutspå defekter i ändringar av nya mjukvaruversioner.
Subjects/Keywords: Just-in-Time Software Defect Prediction; Attention Mechanism; Convolutional Neural Network; Feature Extraction; Just-in-Time Software Defect Prediction; Attention Mechanism; Convolutional Neural Network; Feature Extraction; Computer and Information Sciences; Data- och informationsvetenskap
Record Details
Similar Records
Cite
Share »
Record Details
Similar Records
Cite
« Share
❌
APA ·
Chicago ·
MLA ·
Vancouver ·
CSE |
Export
to Zotero / EndNote / Reference
Manager
APA (6th Edition):
Navarro, A. Y. I. (2020). Evaluation of Attention Mechanisms for Just-In-Time Software Defect Prediction. (Thesis). KTH. Retrieved from http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-288724
Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation
Chicago Manual of Style (16th Edition):
Navarro, Abgeiba Yaroslava Isunza. “Evaluation of Attention Mechanisms for Just-In-Time Software Defect Prediction.” 2020. Thesis, KTH. Accessed March 01, 2021.
http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-288724.
Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation
MLA Handbook (7th Edition):
Navarro, Abgeiba Yaroslava Isunza. “Evaluation of Attention Mechanisms for Just-In-Time Software Defect Prediction.” 2020. Web. 01 Mar 2021.
Vancouver:
Navarro AYI. Evaluation of Attention Mechanisms for Just-In-Time Software Defect Prediction. [Internet] [Thesis]. KTH; 2020. [cited 2021 Mar 01].
Available from: http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-288724.
Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation
Council of Science Editors:
Navarro AYI. Evaluation of Attention Mechanisms for Just-In-Time Software Defect Prediction. [Thesis]. KTH; 2020. Available from: http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-288724
Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation
Delft University of Technology
11.
Tomy, Abhishek (author).
Predicting pedestrian path using optical flow as context cue.
Degree: 2020, Delft University of Technology
URL: http://resolver.tudelft.nl/uuid:1212737e-7b76-4339-94f8-722f443b4764
► A human driver can gauge the intention and signals given by other road users indicative of their future behaviour. The intentions and signals are identified…
(more)
▼ A human driver can gauge the intention and signals given by other road users indicative of their future behaviour. The intentions and signals are identified by looking at the cues originating from vulnerable road users or their surroundings (hand signals, head orientation, posture, traffic signals, distance to curb, etc.). Taking all these cues into account by creating a separate detector for each is an extremely difficult task. Instead, this MSc Thesis will explore the possibility of using a generic contextual cue in optical flow originating from a pedestrian with deep learning methods to improve the path prediction in a naturalistic driving scenario. The contribution of this work is to examine multiple ways to extract relevant information from the optical flow and also explore the possibility of using the entire the high-dimensional optical flow using convolutions and soft-
attention to help identify relevant pixels for the prediction task. This work elaborates on the extraction and processing of optical flow features. It proposes 2 Recurrent neural networks (RNN) based model: one to work with the histogram of optical flow features and the other one to take in the dense optical flow directly. Also, visualization of the soft-
attention weights is done to add a step that helps in the interpretability of the RNN model incorporating dense optical flow. From the experimental results, optical flow features have shown significant improvements in terms of predicting probabilistic confidence for tracks with some changes in their motion mode. It was seen that the convolution-
attention RNN model was able to work with dense optical flow features and position of pedestrians as input to obtain better results among all the combinations of features and models compared in this work.
Advisors/Committee Members: Gavrila, D. (mentor), Pool, E.A.I. (mentor), Ferranti, L. (graduation committee), Delft University of Technology (degree granting institution).
Subjects/Keywords: path prediction; Pedestrian; Pedestrian Crossing Behaviour; Recurrent Neural Network; Attention Mechanism; Optical flow; Vulnerable Road User; deep learning; Intelligent Vehicles; autonomous driving
Record Details
Similar Records
Cite
Share »
Record Details
Similar Records
Cite
« Share
❌
APA ·
Chicago ·
MLA ·
Vancouver ·
CSE |
Export
to Zotero / EndNote / Reference
Manager
APA (6th Edition):
Tomy, A. (. (2020). Predicting pedestrian path using optical flow as context cue. (Masters Thesis). Delft University of Technology. Retrieved from http://resolver.tudelft.nl/uuid:1212737e-7b76-4339-94f8-722f443b4764
Chicago Manual of Style (16th Edition):
Tomy, Abhishek (author). “Predicting pedestrian path using optical flow as context cue.” 2020. Masters Thesis, Delft University of Technology. Accessed March 01, 2021.
http://resolver.tudelft.nl/uuid:1212737e-7b76-4339-94f8-722f443b4764.
MLA Handbook (7th Edition):
Tomy, Abhishek (author). “Predicting pedestrian path using optical flow as context cue.” 2020. Web. 01 Mar 2021.
Vancouver:
Tomy A(. Predicting pedestrian path using optical flow as context cue. [Internet] [Masters thesis]. Delft University of Technology; 2020. [cited 2021 Mar 01].
Available from: http://resolver.tudelft.nl/uuid:1212737e-7b76-4339-94f8-722f443b4764.
Council of Science Editors:
Tomy A(. Predicting pedestrian path using optical flow as context cue. [Masters Thesis]. Delft University of Technology; 2020. Available from: http://resolver.tudelft.nl/uuid:1212737e-7b76-4339-94f8-722f443b4764
University of Sydney
12.
Shi, Fangzhou.
Towards Molecule Generation with Heterogeneous States via Reinforcement Learning
.
Degree: University of Sydney
URL: http://hdl.handle.net/2123/22335
► De novo molecular design and generation are frequently prescribed in the field of chemistry and biology, for it plays a critical role in maintaining the…
(more)
▼ De novo molecular design and generation are frequently prescribed in the field of chemistry and biology, for it plays a critical role in maintaining the prosperity of the chemical industry and benefiting the drug discovery. Nowadays, many significant problems in this field are based on the philosophy of designing molecular structures towards specific desired properties. This research is very meaningful in both medical and AI fields, which can benefits novel drug discovery for some diseases. However, It remains a challenging task due to the large size of chemical space. In recent years, reinforcement learning-based methods leverage graphs to represent molecules and generate molecules as a decision making process. However, this vanilla graph representation may neglect the intrinsic context information with molecules and limits the generation performance accordingly. In this paper, we propose to augment the original graph states with the SMILES context vectors. As a result, SMILES representations are easily processed by a simple language model such that the general semantic features of a molecule can be extracted; and the graph representations perform better in handling the topology relationship of each atom. Moreover, we propose a framework that combines supervised learning and reinforcement learning algorithm to take a solid consideration of these two heterogeneous state representations of a molecule, which can fuse the information from both of them and extract more comprehensive features so that more sophisticated decisions can be made by the policy network. Our model also introduces two attention mechanisms, i.e., action-attention, and graph-attention, to further improve the performance. We conduct our experiments on a practical dataset, ZINC, and the experiment results demonstrate that our framework can outperform other baselines in the learning performance of molecule generation and chemical property optimization.
Subjects/Keywords: molecule generation;
graph representation;
property optimization;
SMILES representation;
attention mechanism;
reinforcement learning
Record Details
Similar Records
Cite
Share »
Record Details
Similar Records
Cite
« Share
❌
APA ·
Chicago ·
MLA ·
Vancouver ·
CSE |
Export
to Zotero / EndNote / Reference
Manager
APA (6th Edition):
Shi, F. (n.d.). Towards Molecule Generation with Heterogeneous States via Reinforcement Learning
. (Thesis). University of Sydney. Retrieved from http://hdl.handle.net/2123/22335
Note: this citation may be lacking information needed for this citation format:
No year of publication.
Not specified: Masters Thesis or Doctoral Dissertation
Chicago Manual of Style (16th Edition):
Shi, Fangzhou. “Towards Molecule Generation with Heterogeneous States via Reinforcement Learning
.” Thesis, University of Sydney. Accessed March 01, 2021.
http://hdl.handle.net/2123/22335.
Note: this citation may be lacking information needed for this citation format:
No year of publication.
Not specified: Masters Thesis or Doctoral Dissertation
MLA Handbook (7th Edition):
Shi, Fangzhou. “Towards Molecule Generation with Heterogeneous States via Reinforcement Learning
.” Web. 01 Mar 2021.
Note: this citation may be lacking information needed for this citation format:
No year of publication.
Vancouver:
Shi F. Towards Molecule Generation with Heterogeneous States via Reinforcement Learning
. [Internet] [Thesis]. University of Sydney; [cited 2021 Mar 01].
Available from: http://hdl.handle.net/2123/22335.
Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation
No year of publication.
Council of Science Editors:
Shi F. Towards Molecule Generation with Heterogeneous States via Reinforcement Learning
. [Thesis]. University of Sydney; Available from: http://hdl.handle.net/2123/22335
Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation
No year of publication.
13.
Costa, Jessica Sophia.
Metilfenidato.
Degree: 2017, Universidade Fernando Pessoa
URL: http://www.rcaap.pt/detail.jsp?id=oai:bdigital.ufp.pt:10284/5979
► O metilfenidato é o psicoestimulante mais empregue no tratamento adjuvante da perturbação de hiperatividade com défice de atenção, uma perturbação de neurodesenvolvimento com grande impacto…
(more)
▼ O metilfenidato é o psicoestimulante mais empregue no tratamento adjuvante da perturbação de hiperatividade com défice de atenção, uma perturbação de neurodesenvolvimento com grande impacto comportamental e social que afeta cerca de 5% das crianças em idade escolar e 2,5% dos adultos a nível mundial. É também utlizado no tratamento da narcolepsia.
Está indicado no tratamento de crianças com idade igual ou superior a 6 anos, adolescentes e adultos que apresentem sintomas severos ou moderados e que não tenham respondido adequadamente ao tratamento psicológico. O seu efeito terapêutico, tem melhorado o quotidiano de muitos pacientes, permitindo-lhes executar as suas tarefas diárias de forma mais tranquila e eficaz.
Nos últimos anos o consumo global deste fármaco tem aumentando significativamente. Em Portugal, a avaliação do número de embalagens de medicamento contendo metilfenidato dispensadas entre 2003 e 2014 denota este consistente aumento e indica terem sido dispensadas em 2014 mais de 276000 embalagens.
No entanto, a utilização contemporânea de metilfenidato ultrapassa o seu uso terapêutico. Os efeitos psicoestimulantes sobre o sistema nervoso central, levaram ao seu consumo ilícito para fins recreativos e como forma de potenciar os desempenhos escolares e laborais. Neste último caso as populações alvo são estudantes, trabalhadores (designadamente motoristas, executivos, profissionais de saúde) e jogadores que procuram melhorar o seu rendimento, aumentando o estado de vigília e os níveis de concentração. Esta utilização de natureza abusiva é cada vez mais acentuada e constitui uma pr eocupação crescente para a comunidade representando um grave problema de saúde pública que tem sido alvo de preocupação pelos profissionais de saúde.
Acentua-se, por isso, a necessidade de monitorização das concentrações plasmáticas do fármaco. Nos últimos anos, têm sido desenvolvidas metodologias analíticas que permitem a enantioseparação e quantificação de metilfenidato e do seu principal metabolito, o ácido ritalínico, utilizando cromatografia gasosa, cromatografia líquida de elevada eficiência ou técnicas de eletroforese capilar acopladas a uma variedade de detetores.
Methylphenidate is the most commonly psychostimulant used in the adjuvant treatment of attention deficit hyperactivity disorder, a neurodevelopmental disorder with major behavioral and social impact affecting about 5% of school-aged children and 2.5% of adults worldwide. It is also used in the treatment of narcolepsy.
It's recommended for the treatment of children older than 6 years, adolescents and adults who present severe to moderate symptoms and did not respond adequately to psychological treatment. Its therapeutic effect has improved the daily routines of many patients, enabling them to perform their daily tasks in a more orderly and effective manner.
In recent years the overall consumption of this drug has increased significantly. In Portugal, the assessment of the number of medical products containing methylphenidate dispensed between 2003 and 2014 shows…
Advisors/Committee Members: Souto, Renata, Catarino, Rita.
Subjects/Keywords: Metilfenidato; Perturbação de hiperatividade com défice de atenção; Narcolepsia; Mecanismo de ação; Quantificação; Técnicas analíticas; Methylphenidate; Attention deficit hyperactivity disorder; Narcolepsy; Mechanism of action; Quantification; Analytical techniques; Domínio/Área Científica::Ciências Médicas::Medicina Básica
Record Details
Similar Records
Cite
Share »
Record Details
Similar Records
Cite
« Share
❌
APA ·
Chicago ·
MLA ·
Vancouver ·
CSE |
Export
to Zotero / EndNote / Reference
Manager
APA (6th Edition):
Costa, J. S. (2017). Metilfenidato. (Thesis). Universidade Fernando Pessoa. Retrieved from http://www.rcaap.pt/detail.jsp?id=oai:bdigital.ufp.pt:10284/5979
Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation
Chicago Manual of Style (16th Edition):
Costa, Jessica Sophia. “Metilfenidato.” 2017. Thesis, Universidade Fernando Pessoa. Accessed March 01, 2021.
http://www.rcaap.pt/detail.jsp?id=oai:bdigital.ufp.pt:10284/5979.
Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation
MLA Handbook (7th Edition):
Costa, Jessica Sophia. “Metilfenidato.” 2017. Web. 01 Mar 2021.
Vancouver:
Costa JS. Metilfenidato. [Internet] [Thesis]. Universidade Fernando Pessoa; 2017. [cited 2021 Mar 01].
Available from: http://www.rcaap.pt/detail.jsp?id=oai:bdigital.ufp.pt:10284/5979.
Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation
Council of Science Editors:
Costa JS. Metilfenidato. [Thesis]. Universidade Fernando Pessoa; 2017. Available from: http://www.rcaap.pt/detail.jsp?id=oai:bdigital.ufp.pt:10284/5979
Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation
Pontifical Catholic University of Rio de Janeiro
14.
MAURO PINHEIRO RODRIGUES.
[en] INTERACTION DESIGN AND PERVASIVE COMPUTING: A STUDY OF
ATTENTIONAL MECHANISMS AND AMBIENT INFORMATION SYSTEMS.
Degree: 2013, Pontifical Catholic University of Rio de Janeiro
URL: http://www.maxwell.vrac.puc-rio.br/Busca_etds.php?strSecao=resultado&nrSeq=21718
► [pt] A tese investiga os diferentes mecanismos atencionais envolvidos na utilização de sistemas de informação ambiente. Para tanto, conceitua e delimita a computação pervasiva –…
(more)
▼ [pt] A tese investiga os diferentes mecanismos
atencionais envolvidos na utilização de sistemas de informação
ambiente. Para tanto, conceitua e delimita a computação pervasiva –
a partir da qual componentes computadorizados passam a compor o
ambiente e os objetos do cotidiano –, e discute as conseqüências do
uso desta tecnologia, no que se refere aos impactos sociais,
implicações ambientais, questões de segurança e privacidade,
destacando o papel do design nessa problemática. Apresenta a
evolução do design de interação, explicitando sua relação com o
projeto de mídias interativas. Propõe ampliar o campo de atuação do
design de interação, considerando que a tela do computador deixou
de ser a principal interface com o ambiente digital, e que o
projeto da interação no contexto da computação pervasiva exige uma
abordagem sistêmica. Apresenta o conceito de tecnologia sem
estresse (calm technology), de Weiser e Brown (1996), e aprofunda a
discussão iniciada por estes autores sobre a necessidade de se
projetar sistemas de informação que atuem na periferia de nossa
atenção, com base nos estudos da Psicologia Cognitiva sobre
mecanismos atencionais. Destaca os sistemas de informação ambiente
como aqueles que mais se apropriam da idéia de apresentar
informações sem exigir o foco de nossa atenção. Analisa doze
sistemas de informação ambiente, investigando o modo como envolvem
os mecanismos atencionais. Conclui que a definição original de
Weiser e Brown (1996) não é suficiente para descrever a miríade de
processos envolvidos com a captação da atenção, e aponta linhas
mestras para o design de sistemas de informação ambiente, de
maneira a considerar a dinâmica entre os diferentes mecanismos
atencionais, o contexto de uso, o grau de engajamento do usuário, a
influência da memória e a capacidade de habituação aos sistemas de
informação.
[en] This thesis investigates the different
attentional mechanisms involved when using ambient information
systems. To that end, it defines and delimits pervasive computing –
when computational resources are embedded into the environment and
in everyday objects – and discusses the consequences of this
technology, regarding the social impacts, environmental
implications, security and privacy issues, highlighting the role of
design on this matter. It presents the evolution of interaction
design, emphasizing its relationship with the design of interactive
media. It proposes to broaden the interaction design field,
considering that the computer screen is no longer the primary
interface with the digital environment, and that interaction design
requires a systemic approach in the context of pervasive computing.
It introduces Weiser and Brown s (1996) concept of calm technology,
and deepens the discussion initiated by these authors about the
need for designing information systems that act on the periphery of
our attention, based on Cognitive psychology studies about
attentional mechanisms. It highlights ambient information systems
as those which have more properly embraced the idea of presenting…
Advisors/Committee Members: REJANE SPITZ.
Subjects/Keywords: [pt] PERCEPCAO; [en] PERCEPTION; [pt] ATENCAO; [en] ATTENTION; [pt] COMPUTACAO UBIQUA; [en] UBIQUITOUS COMPUTING; [pt] COMPUTACAO PERVASIVA; [en] PERVASIVE COMPUTING; [pt] MECANISMO ATENCIONAL; [en] ATTENTIONAL MECHANISM; [pt] SISTEMAS DE INFORMACAO AMBIENTE; [en] AMBIENT INFORMATION SYSTEMS; [pt] DESIGN DE INTERACAO; [en] INTERACTION DESIGN
Record Details
Similar Records
Cite
Share »
Record Details
Similar Records
Cite
« Share
❌
APA ·
Chicago ·
MLA ·
Vancouver ·
CSE |
Export
to Zotero / EndNote / Reference
Manager
APA (6th Edition):
RODRIGUES, M. P. (2013). [en] INTERACTION DESIGN AND PERVASIVE COMPUTING: A STUDY OF
ATTENTIONAL MECHANISMS AND AMBIENT INFORMATION SYSTEMS. (Thesis). Pontifical Catholic University of Rio de Janeiro. Retrieved from http://www.maxwell.vrac.puc-rio.br/Busca_etds.php?strSecao=resultado&nrSeq=21718
Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation
Chicago Manual of Style (16th Edition):
RODRIGUES, MAURO PINHEIRO. “[en] INTERACTION DESIGN AND PERVASIVE COMPUTING: A STUDY OF
ATTENTIONAL MECHANISMS AND AMBIENT INFORMATION SYSTEMS.” 2013. Thesis, Pontifical Catholic University of Rio de Janeiro. Accessed March 01, 2021.
http://www.maxwell.vrac.puc-rio.br/Busca_etds.php?strSecao=resultado&nrSeq=21718.
Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation
MLA Handbook (7th Edition):
RODRIGUES, MAURO PINHEIRO. “[en] INTERACTION DESIGN AND PERVASIVE COMPUTING: A STUDY OF
ATTENTIONAL MECHANISMS AND AMBIENT INFORMATION SYSTEMS.” 2013. Web. 01 Mar 2021.
Vancouver:
RODRIGUES MP. [en] INTERACTION DESIGN AND PERVASIVE COMPUTING: A STUDY OF
ATTENTIONAL MECHANISMS AND AMBIENT INFORMATION SYSTEMS. [Internet] [Thesis]. Pontifical Catholic University of Rio de Janeiro; 2013. [cited 2021 Mar 01].
Available from: http://www.maxwell.vrac.puc-rio.br/Busca_etds.php?strSecao=resultado&nrSeq=21718.
Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation
Council of Science Editors:
RODRIGUES MP. [en] INTERACTION DESIGN AND PERVASIVE COMPUTING: A STUDY OF
ATTENTIONAL MECHANISMS AND AMBIENT INFORMATION SYSTEMS. [Thesis]. Pontifical Catholic University of Rio de Janeiro; 2013. Available from: http://www.maxwell.vrac.puc-rio.br/Busca_etds.php?strSecao=resultado&nrSeq=21718
Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation
Linköping University
15.
Govindarajan, Hariprasath.
Self-Supervised Representation Learning for Content Based Image Retrieval.
Degree: The Division of Statistics and Machine Learning, 2020, Linköping University
URL: http://urn.kb.se/resolve?urn=urn:nbn:se:liu:diva-166223
► Automotive technologies and fully autonomous driving have seen a tremendous growth in recent times and have benefitted from extensive deep learning research. State-of-the-art deep…
(more)
▼ Automotive technologies and fully autonomous driving have seen a tremendous growth in recent times and have benefitted from extensive deep learning research. State-of-the-art deep learning methods are largely supervised and require labelled data for training. However, the annotation process for image data is time-consuming and costly in terms of human efforts. It is of interest to find informative samples for labelling by Content Based Image Retrieval (CBIR). Generally, a CBIR method takes a query image as input and returns a set of images that are semantically similar to the query image. The image retrieval is achieved by transforming images to feature representations in a latent space, where it is possible to reason about image similarity in terms of image content. In this thesis, a self-supervised method is developed to learn feature representations of road scenes images. The self-supervised method learns feature representations for images by adapting intermediate convolutional features from an existing deep Convolutional Neural Network (CNN). A contrastive approach based on Noise Contrastive Estimation (NCE) is used to train the feature learning model. For complex images like road scenes where mutiple image aspects can occur simultaneously, it is important to embed all the salient image aspects in the feature representation. To achieve this, the output feature representation is obtained as an ensemble of feature embeddings which are learned by focusing on different image aspects. An attention mechanism is incorporated to encourage each ensemble member to focus on different image aspects. For comparison, a self-supervised model without attention is considered and a simple dimensionality reduction approach using SVD is treated as the baseline. The methods are evaluated on nine different evaluation datasets using CBIR performance metrics. The datasets correspond to different image aspects and concern the images at different spatial levels - global, semi-global and local. The feature representations learned by self-supervised methods are shown to perform better than the SVD approach. Taking into account that no labelled data is required for training, learning representations for road scenes images using self-supervised methods appear to be a promising direction. Usage of multiple query images to emphasize a query intention is investigated and a clear improvement in CBIR performance is observed. It is inconclusive whether the addition of an attentive mechanism impacts CBIR performance. The attention method shows some positive signs based on qualitative analysis and also performs better than other methods for one of the evaluation datasets containing a local aspect. This method for learning feature representations is promising but requires further research involving more diverse and complex image aspects.
Subjects/Keywords: Content Based Image Retrieval; CBIR; Representation Learning; Self Supervised Learning; Unsupervised Learning; Attention Mechanism; Noise Contrastive Estimation; Autonomous Driving; Computer and Information Sciences; Data- och informationsvetenskap; Probability Theory and Statistics; Sannolikhetsteori och statistik
Record Details
Similar Records
Cite
Share »
Record Details
Similar Records
Cite
« Share
❌
APA ·
Chicago ·
MLA ·
Vancouver ·
CSE |
Export
to Zotero / EndNote / Reference
Manager
APA (6th Edition):
Govindarajan, H. (2020). Self-Supervised Representation Learning for Content Based Image Retrieval. (Thesis). Linköping University. Retrieved from http://urn.kb.se/resolve?urn=urn:nbn:se:liu:diva-166223
Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation
Chicago Manual of Style (16th Edition):
Govindarajan, Hariprasath. “Self-Supervised Representation Learning for Content Based Image Retrieval.” 2020. Thesis, Linköping University. Accessed March 01, 2021.
http://urn.kb.se/resolve?urn=urn:nbn:se:liu:diva-166223.
Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation
MLA Handbook (7th Edition):
Govindarajan, Hariprasath. “Self-Supervised Representation Learning for Content Based Image Retrieval.” 2020. Web. 01 Mar 2021.
Vancouver:
Govindarajan H. Self-Supervised Representation Learning for Content Based Image Retrieval. [Internet] [Thesis]. Linköping University; 2020. [cited 2021 Mar 01].
Available from: http://urn.kb.se/resolve?urn=urn:nbn:se:liu:diva-166223.
Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation
Council of Science Editors:
Govindarajan H. Self-Supervised Representation Learning for Content Based Image Retrieval. [Thesis]. Linköping University; 2020. Available from: http://urn.kb.se/resolve?urn=urn:nbn:se:liu:diva-166223
Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation
16.
Çinar, Yagmur Gizem.
Prédiction de séquences basée sur des réseaux de neurones récurrents dans le contexte des séries temporelles et des sessions de recherche d'information : Sequence Prediction using Recurrent Neural Networks in the Context of Time Series and Information Retrieval Search Sessions.
Degree: Docteur es, Informatique, 2019, Université Grenoble Alpes (ComUE)
URL: http://www.theses.fr/2019GREAM079
► Cette thèse examine les défis de la prédiction de séquence dans différents scénarios, tels que la prédiction de séquence à l'aide de réseaux de neurones…
(more)
▼ Cette thèse examine les défis de la prédiction de séquence dans différents scénarios, tels que la prédiction de séquence à l'aide de réseaux de neurones récurrents (RNN) dans le contexte des séries temporelles et des sessions de recherche d'informations (RI). Prédire les valeurs inconnues suivant certaines valeurs précédemment observées est appelée prédiction de séquence. Elle est largement applicable à de nombreux domaines où un comportement séquentiel est observé dans les données. Dans cette étude, nous nous concentrons sur deux tâches de prédiction de séquences: la prévision de séries temporelles et la prédiction de la requête suivante dans une session de recherche d'informations.Les séries temporelles comprennent souvent des pseudo-périodes, c'est-à-dire des intervalles de temps avec une forte corrélation entre les valeurs des séries temporelles. Les changements saisonniers dans les séries temporelles météorologiques ou la consommation d'électricité le jour et la nuit sont quelques exemples de pseudo-périodes. Dans un scénario de prévision, les pseudo-périodes correspondent à la différence entre les positions de la sortie prévue et les entrées spécifiques. Afin de capturer des périodes dans des RNN, une mémoire de la séquence d'entrée est requise. Les RNN séquence à séquence (avec mécanisme d'
attention) réutilisent des (représentations des) valeurs d'entrée spécifiques pour prédire les valeurs de sortie. Les RNN séquence à séquence avec un mécanisme d'
attention semblent convenir à la capture de périodes. Ainsi, nous explorons d’abord la capacité d’un mécanisme d’
attention dans ce contexte. Cependant, selon notre analyse initiale, un mécanisme d’
attention standard ne permet pas de capturer les périodes. Par conséquent, nous proposons un modèle RNN d’
attention basé sur le contenu et sensible à la période. Ce modèle étend les RNN séquence à séquence de l'état de l'art avec un mécanisme d’
attention. Il vise à capturer les périodes dans une série temporelle avec ou sans valeurs manquantes. Nos résultats expérimentaux avec des RNN contenant un mécanisme d'
attention basé sur le contenu et sensible à la période montrent une amélioration significative des performances de prévision des séries temporelles univariées et multivariées sur plusieurs ensembles de données disponibles publiquement.La prédiction de la requête suivante est un autre défi de la prédiction de séquence. La prédiction de la requête suivante aide les utilisateurs à désambiguïser leur requête, à explorer différents aspects de leur besoin en information ou à former une requête précise et succincte qui permet d’optimiser les performances de la recherche. Une session de recherche est dynamique et les besoins en informations d'un utilisateur peuvent changer au cours d'une session de recherche à la suite des interactions de recherche. De plus, les interactions d'un utilisateur avec un moteur de recherche influencent les reformulations de requêtes de l'utilisateur. Considérant cette influence sur les formulations de requête, nous analysons d’abord l’origine des…
Advisors/Committee Members: Gaussier, Éric (thesis director).
Subjects/Keywords: Apprentissage automatique; Recherche d'informations; Réseaux de neurones récurrents; Mécanisme d’attention; Prévision de séries temporelles; Prédiction de la requête suivante; Machine learning; Information retrieval; Recurrent neural networks; Attention mechanism; Time series forecasting; Next query prediction; 004
Record Details
Similar Records
Cite
Share »
Record Details
Similar Records
Cite
« Share
❌
APA ·
Chicago ·
MLA ·
Vancouver ·
CSE |
Export
to Zotero / EndNote / Reference
Manager
APA (6th Edition):
Çinar, Y. G. (2019). Prédiction de séquences basée sur des réseaux de neurones récurrents dans le contexte des séries temporelles et des sessions de recherche d'information : Sequence Prediction using Recurrent Neural Networks in the Context of Time Series and Information Retrieval Search Sessions. (Doctoral Dissertation). Université Grenoble Alpes (ComUE). Retrieved from http://www.theses.fr/2019GREAM079
Chicago Manual of Style (16th Edition):
Çinar, Yagmur Gizem. “Prédiction de séquences basée sur des réseaux de neurones récurrents dans le contexte des séries temporelles et des sessions de recherche d'information : Sequence Prediction using Recurrent Neural Networks in the Context of Time Series and Information Retrieval Search Sessions.” 2019. Doctoral Dissertation, Université Grenoble Alpes (ComUE). Accessed March 01, 2021.
http://www.theses.fr/2019GREAM079.
MLA Handbook (7th Edition):
Çinar, Yagmur Gizem. “Prédiction de séquences basée sur des réseaux de neurones récurrents dans le contexte des séries temporelles et des sessions de recherche d'information : Sequence Prediction using Recurrent Neural Networks in the Context of Time Series and Information Retrieval Search Sessions.” 2019. Web. 01 Mar 2021.
Vancouver:
Çinar YG. Prédiction de séquences basée sur des réseaux de neurones récurrents dans le contexte des séries temporelles et des sessions de recherche d'information : Sequence Prediction using Recurrent Neural Networks in the Context of Time Series and Information Retrieval Search Sessions. [Internet] [Doctoral dissertation]. Université Grenoble Alpes (ComUE); 2019. [cited 2021 Mar 01].
Available from: http://www.theses.fr/2019GREAM079.
Council of Science Editors:
Çinar YG. Prédiction de séquences basée sur des réseaux de neurones récurrents dans le contexte des séries temporelles et des sessions de recherche d'information : Sequence Prediction using Recurrent Neural Networks in the Context of Time Series and Information Retrieval Search Sessions. [Doctoral Dissertation]. Université Grenoble Alpes (ComUE); 2019. Available from: http://www.theses.fr/2019GREAM079
17.
Simonnet, Edwin.
Réseaux de neurones profonds appliqués à la compréhension de la parole : Deep learning applied to spoken langage understanding.
Degree: Docteur es, Informatique, 2019, Le Mans
URL: http://www.theses.fr/2019LEMA1006
► Cette thèse s'inscrit dans le cadre de l'émergence de l'apprentissage profond et aborde la compréhension de la parole assimilée à l'extraction et à la représentation…
(more)
▼ Cette thèse s'inscrit dans le cadre de l'émergence de l'apprentissage profond et aborde la compréhension de la parole assimilée à l'extraction et à la représentation automatique du sens contenu dans les mots d'une phrase parlée. Nous étudions une tâche d'étiquetage en concepts sémantiques dans un contexte de dialogue oral évaluée sur le corpus français MEDIA. Depuis une dizaine d'années, les modèles neuronaux prennent l'ascendant dans de nombreuses tâches de traitement du langage naturel grâce à des avancées algorithmiques ou à la mise à disposition d'outils de calcul puissants comme les processeurs graphiques. De nombreux obstacles rendent la compréhension complexe, comme l'interprétation difficile des transcriptions automatiques de la parole étant donné que de nombreuses erreurs sont introduites par le processus de reconnaissance automatique en amont du module de compréhension. Nous présentons un état de l'art décrivant la compréhension de la parole puis les méthodes d'apprentissage automatique supervisé pour la résoudre en commençant par des systèmes classiques pour finir avec des techniques d'apprentissage profond. Les contributions sont ensuite exposées suivant trois axes. Premièrement, nous développons une architecture neuronale efficace consistant en un réseau récurent bidirectionnel encodeur-décodeur avec mécanisme d’attention. Puis nous abordons la gestion des erreurs de reconnaissance automatique et des solutions pour limiter leur impact sur nos performances. Enfin, nous envisageons une désambiguïsation de la tâche de compréhension permettant de rendre notre système plus performant.
This thesis is a part of the emergence of deep learning and focuses on spoken language understanding assimilated to the automatic extraction and representation of the meaning supported by the words in a spoken utterance. We study a semantic concept tagging task used in a spoken dialogue system and evaluated with the French corpus MEDIA. For the past decade, neural models have emerged in many natural language processing tasks through algorithmic advances or powerful computing tools such as graphics processors. Many obstacles make the understanding task complex, such as the difficult interpretation of automatic speech transcriptions, as many errors are introduced by the automatic recognition process upstream of the comprehension module. We present a state of the art describing spoken language understanding and then supervised automatic learning methods to solve it, starting with classical systems and finishing with deep learning techniques. The contributions are then presented along three axes. First, we develop an efficient neural architecture consisting of a bidirectional recurrent network encoder-decoder with attention mechanism. Then we study the management of automatic recognition errors and solutions to limit their impact on our performances. Finally, we envisage a disambiguation of the comprehension task making the systems more efficient.
Advisors/Committee Members: Estève, Yannick (thesis director), Camelin, Nathalie (thesis director).
Subjects/Keywords: Compréhension de la parole; Corpus MEDIA; Étiquetage en concept sémantiques; Réseaux de neurones profonds; Mécanisme d'attention; Erreurs de reconnaissance automatique; Simulation d'erreurs de reconnaissance; Désambiguïsation de la compréhension; Spoken Language Understanding; MEDIA corpus; Semantic concept tagging; Deep learning; Attention mechanism; Automatic speech recognition errors; Simulation of recognition errors; Disambiguation of understanding; 006.454
Record Details
Similar Records
Cite
Share »
Record Details
Similar Records
Cite
« Share
❌
APA ·
Chicago ·
MLA ·
Vancouver ·
CSE |
Export
to Zotero / EndNote / Reference
Manager
APA (6th Edition):
Simonnet, E. (2019). Réseaux de neurones profonds appliqués à la compréhension de la parole : Deep learning applied to spoken langage understanding. (Doctoral Dissertation). Le Mans. Retrieved from http://www.theses.fr/2019LEMA1006
Chicago Manual of Style (16th Edition):
Simonnet, Edwin. “Réseaux de neurones profonds appliqués à la compréhension de la parole : Deep learning applied to spoken langage understanding.” 2019. Doctoral Dissertation, Le Mans. Accessed March 01, 2021.
http://www.theses.fr/2019LEMA1006.
MLA Handbook (7th Edition):
Simonnet, Edwin. “Réseaux de neurones profonds appliqués à la compréhension de la parole : Deep learning applied to spoken langage understanding.” 2019. Web. 01 Mar 2021.
Vancouver:
Simonnet E. Réseaux de neurones profonds appliqués à la compréhension de la parole : Deep learning applied to spoken langage understanding. [Internet] [Doctoral dissertation]. Le Mans; 2019. [cited 2021 Mar 01].
Available from: http://www.theses.fr/2019LEMA1006.
Council of Science Editors:
Simonnet E. Réseaux de neurones profonds appliqués à la compréhension de la parole : Deep learning applied to spoken langage understanding. [Doctoral Dissertation]. Le Mans; 2019. Available from: http://www.theses.fr/2019LEMA1006
. 
Open Access Theses and Dissertations