A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
Filters
FR-Train: A Mutual Information-Based Approach to Fair and Robust Training
[article]
2020
arXiv
pre-print
Preserved Fulltext
Other Versions
We provide a mutual information-based interpretation of an existing adversarial training-based fairness-only method, and apply this idea to architect an additional discriminator that can identify poisoned ...
To address this problem, we propose FR-Train, which holistically performs fair and robust model training. ...
Robustness The robustness discriminator ensures robust training by using mutual information to distinguish examples and predictions from a clean validation set. ...
arXiv:2002.10234v2
fatcat:5lqxsxfkhrcivfl5imgqfckjee
A Sociotechnical View of Algorithmic Fairness
[article]
2021
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2021; you can also visit the original URL.
The file type is application/pdf.
We call for and undertake a holistic approach to AF. A sociotechnical perspective on algorithmic fairness can yield holistic solutions to systemic biases and discrimination. ...
However, based on a state-of-the-art literature review, we argue that fairness is an inherently social concept and that technologies for algorithmic fairness should therefore be approached through a sociotechnical ...
., Suh, C. (2020).FR-Train: A Mutual Information-Based Approach to Fair and Robust Training. ...
arXiv:2110.09253v1
fatcat:sohgqzk7dfbnjlevmhznreui5e
Tilted Empirical Risk Minimization
[article]
2021
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2021; you can also visit the original URL.
The file type is application/pdf.
Other Versions
properties that can benefit generalization; and can be viewed as a smooth approximation to a superquantile method. ...
We provide several interpretations of the resulting framework: We show that TERM can increase or decrease the influence of outliers, respectively, to enable fairness or robustness; has variance-reduction ...
ACKNOWLEDGEMENTS We are grateful to Arun Sai Suggala and Adarsh Prasad (CMU) for their helpful comments on robust regression; to Zhiguang Wang, Dario Garcia Garcia, Alborz Geramifard, and other members ...
arXiv:2007.01162v2
fatcat:bkwwhpdesvdabiebaydgh3r4ry
Face Image Quality Assessment: A Literature Survey
2022
ACM Computing Surveys
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2022; you can also visit the original URL.
The file type is application/pdf.
A trend towards deep learning based methods is observed, including notable conceptual differences among the recent approaches, such as the integration of quality assessment into face recognition models ...
Open issues and challenges are pointed out, i.a. highlighting the importance of comparability for algorithm evaluations, and the challenge for future work to create deep learning approaches that are interpretable ...
MSM (Mutual Subspace Method) based on either LBP or HOG features. ...
doi:10.1145/3507901
fatcat:xvs67qamgbbdtjydzekinnq62u
Meta Balanced Network for Fair Face Recognition
2021
IEEE Transactions on Pattern Analysis and Machine Intelligence
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2022; you can also visit the original URL.
The file type is application/pdf.
Note that this fulltext copy is not of the "primary" version of this work. The version it corresponds to is:
2022 pre-print arXiv:2205.06548v1 fatcat:mzi7dtksiffgxmjukjy7gveyia
Further, we provide two skin-tone aware training datasets, called BUPT-Globalface dataset and BUPT-Balancedface dataset, to remove bias in training data. ...
To determine the margins, our method optimizes a meta skewness loss on a clean and unbiased meta set and utilizes backward-on-backward automatic differentiation to perform a second order gradient descent ...
in the similar way as other FR training datasets, e.g., VGGface2 [19] and Megaface [65] . ...
doi:10.1109/tpami.2021.3103191
fatcat:s4bq32k4pjfbphgpvzw6jatqfm
Face Image Quality Assessment: A Literature Survey
[article]
2021
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2021; you can also visit the original URL.
The file type is application/pdf.
Other Versions
A trend towards deep learning based methods is observed, including notable conceptual differences among the recent approaches, such as the integration of quality assessment into face recognition models ...
Open issues and challenges are pointed out, i.a. highlighting the importance of comparability for algorithm evaluations, and the challenge for future work to create deep learning approaches that are interpretable ...
This text reflects only the author's views and the Commission is not liable for any use that may be made of the information contained therein. ...
arXiv:2009.01103v3
fatcat:ilhbi4vwgfg3ld2wynty5aqjzq
Fairness-Aware PAC Learning from Corrupted Data
[article]
2021
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2021; you can also visit the original URL.
The file type is application/pdf.
Other Versions
While many approaches have been developed for training fair models from data, little is known about the robustness of these methods to data corruption. ...
Addressing fairness concerns about machine learning models is a crucial step towards their long-term adoption in real-world automated systems. ...
Yuji Roh, Kangwook Lee, Steven Euijong Whang, and Changho Suh. FR-Train: A mutual
information-based approach to fair and robust training. ...
arXiv:2102.06004v2
fatcat:n2pcxaougffulewkxl3uw6mt2q
Where the devil dances: A constructivist grounded theory of resilience in volunteer firefighters
2017
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2022; you can also visit the original URL.
The file type is application/pdf.
The purpose of this programme of study was to construct a theory of resilience in volunteer firefighters, a population that, despite facing intermittent and at times intense work-related stressors, is ...
The CGT asserts that within a volunteer FRS there are a number of concepts that inter-relate to construct resilience: relationships, personal resources, meaning-making, leadership, culture, and knowledge ...
Historically, stress has been conceived of as a medical or pathological response; however, recognizing firefighters experience normal and expected responses to stress at both the micro and ...
doi:10.25316/ir-223
fatcat:tjpyhasmhfg7ni5iikglvhwkmi
Energy Solutions for Smart Cities and Communities Recommendations for Policy Makers from the 58 Pilots of the CONCERTO Initiative
unpublished
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2018; you can also visit the original URL.
The file type is application/pdf.
Apart from the policy recommendations presented here, a technical monitoring data base (TMD) and a comprehensive assessment of the data, supplied by the 58 projects, are the main outputs of CONCERTO Premium ...
The database, as well as project results and links to single projects, are available to the public at www.concerto.eu. ...
The main objective of CONCERTO Premium is to provide a robust data and information base that serves as the foundation for setting future regulatory frameworks and as a support for investors` decision-making ...
fatcat:dgq6aqhb7fhedhk5jyhduwd6zm