Rethinking Common Assumptions to Mitigate Racial Bias in Face Recognition Datasets

Gwilliam, Matthew; Hegde, Srinidhi; Tinubu, Lade; Hanson, Alex

doi:10.1109/iccvw54120.2021.00458

Cited by 25 publications

(11 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A not-expected approach is seen as the researchers claim that training only on one race is not inherently disadvantageous [102]. They demonstrate that by training only using African faces, they achieved less skew across race than by training with a balanced dataset.…”

Section: Network Improvements On the Rfw Datasetmentioning

confidence: 99%

Surveying Racial Bias in Facial Recognition: Balancing Datasets and Algorithmic Enhancements

Sumsion,

Torrie,

Lee

et al. 2024

Electronics

View full text Add to dashboard Cite

Facial recognition systems frequently exhibit high accuracies when evaluated on standard test datasets. However, their performance tends to degrade significantly when confronted with more challenging tests, particularly involving specific racial categories. To measure this inconsistency, many have created racially aware datasets to evaluate facial recognition algorithms. This paper analyzes facial recognition datasets, categorizing them as racially balanced or unbalanced while limiting racially balanced datasets to have each race be represented within five percentage points of all other represented races. We investigate methods to address concerns about racial bias due to uneven datasets by using generative adversarial networks and latent diffusion models to balance the data, and we also assess the impact of these techniques. In an effort to mitigate accuracy discrepancies across different racial groups, we investigate a range of network enhancements in facial recognition performance across human races. These improvements encompass architectural improvements, loss functions, training methods, data modifications, and incorporating additional data. Additionally, we discuss the interrelation of racial and gender bias. Lastly, we outline avenues for future research in this domain.

show abstract

Section: Network Improvements On the Rfw Datasetmentioning

confidence: 99%

Surveying Racial Bias in Facial Recognition: Balancing Datasets and Algorithmic Enhancements

Sumsion,

Torrie,

Lee

et al. 2024

Electronics

View full text Add to dashboard Cite

show abstract

“…An alternative road is to use data augmentation techniques to "rebalance" the dataset [27,45]. However, it was discovered that it is not sufficient for avoiding bias to use an assumed balanced datasets during training [20,49,50] because it is often unclear which features in the data need to be balanced. Approaches for curating or manipulating the dataset require information on the target domain, i.e., one needs to set requirements on the dataset depending on the desired operational context [6,16,22].…”

Section: Related Literaturementioning

confidence: 99%

An investigation of challenges encountered when specifying training data and runtime monitors for safety critical ML applications

Heyn¹,

Knauss²,

Malleswaran³

et al. 2023

Preprint

View full text Add to dashboard Cite

The development and operation of critical software that contains machine learning (ML) models requires diligence and established processes. Especially the training data used during the development of ML models have major influences on the later behaviour of the system. Runtime monitors are used to provide guarantees for that behaviour. [Question / problem] We see major uncertainty in how to specify training data and runtime monitoring for critical ML models and by this specifying the final functionality of the system. In this interview-based study we investigate the underlying challenges for these difficulties. [Principal ideas/results] Based on ten interviews with practitioners who develop ML models for critical applications in the automotive and telecommunication sector, we identified 17 underlying challenges in 6 challenge groups that relate to the challenge of specifying training data and runtime monitoring.[Contribution] The article provides a list of the identified underlying challenges related to the difficulties practitioners experience when specifying training data and runtime monitoring for ML models. Furthermore, interconnection between the challenges were found and based on these connections recommendation proposed to overcome the root causes for the challenges.

show abstract

“…The authors leveraged synthetic data for analysis and showcased how facial pose and facial identity cannot be completely disentangled by deep networks (bias in model training Figure 2(c)). To further study the impact of dataset bias, (Gwilliam et al 2021) analyzed facial recognition performance by training on various imbalanced distributions across race. They observed less biased model predictions after training on a specific subgroup than training on a balanced distribution.…”

Section: Bias In Face Detection and Recognitionmentioning

confidence: 99%

Anatomizing Bias in Facial Analysis

Singh

Majumdar

Mittal

et al. 2022

AAAI

View full text Add to dashboard Cite

Existing facial analysis systems have been shown to yield biased results against certain demographic subgroups. Due to its impact on society, it has become imperative to ensure that these systems do not discriminate based on gender, identity, or skin tone of individuals. This has led to research in the identification and mitigation of bias in AI systems. In this paper, we encapsulate bias detection/estimation and mitigation algorithms for facial analysis. Our main contributions include a systematic review of algorithms proposed for understanding bias, along with a taxonomy and extensive overview of existing bias mitigation algorithms. We also discuss open challenges in the field of biased facial analysis.

show abstract

Rethinking Common Assumptions to Mitigate Racial Bias in Face Recognition Datasets

Cited by 25 publications

References 25 publications

Surveying Racial Bias in Facial Recognition: Balancing Datasets and Algorithmic Enhancements

Surveying Racial Bias in Facial Recognition: Balancing Datasets and Algorithmic Enhancements

An investigation of challenges encountered when specifying training data and runtime monitors for safety critical ML applications

Anatomizing Bias in Facial Analysis

Contact Info

Product

Resources

About