Turning a Blind Eye: Explicit Removal of Biases and Variation from Deep Neural Network Embeddings

Alvi, Mohsan; Zisserman, Andrew; Nellåker, Christoffer

doi:10.48550/arxiv.1809.02169

Cited by 3 publications

(19 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In many computer vision applications, there are some works that seek to introduce fairness into networks and mitigate data bias. These are respectively classified as unbalanced-training [40,10,35,52], attribute suppression [5,32,36,31] and domain adaptation [24,23,49,43]. By learning the underlying latent variables in an entirely unsupervised manner, Debiasing Variational Autoencoder (DB-VAE) [6] re-weighted the importance of certain data points while training.…”

Section: Debiased Algorithmsmentioning

confidence: 99%

“…Recently, with the emergence of deep convolutional neural networks (CNN) [26,42,45,20,21], the performance of face recognition (FR) [48,44,41] is dramatically boosted. However, as its wider and wider application, its potential for unfairness is raising alarm [9,5,1,2]. For instance, Amazons Rekognition Tool incorrectly matched the photos of 28 U.S. congressmen with the faces of criminals, especially the error rate was up to 39% for non-Caucasian people; according to [15], a year-long research investigation across A major driver of bias in face recognition, as well as other AI tasks, is the training data.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Mitigate Bias in Face Recognition using Skewness-Aware Reinforcement Learning

Wang¹,

Deng²

2019

Preprint

View full text Add to dashboard Cite

Racial equality is an important theme of international human rights law, but it has been largely obscured when the overall face recognition accuracy is pursued blindly. More facts indicate racial bias indeed degrades the fairness of recognition system and the error rates on non-Caucasians are usually much higher than Caucasians. To encourage fairness, we introduce the idea of adaptive margin to learn balanced performance for different races based on large margin losses. A reinforcement learning based race balance network (RL-RBN) is proposed. We formulate the process of finding the optimal margins for non-Caucasians as a Markov decision process and employ deep Q-learning to learn policies for an agent to select appropriate margin by approximating the Q-value function. Guided by the agent, the skewness of feature scatter between races can be reduced. Besides, we provide two ethnicity aware training datasets, called BUPT-Globalface and BUPT-Balancedface dataset, which can be utilized to study racial bias from both data and algorithm aspects. Extensive experiments on RFW database show that RL-RBN successfully mitigates racial bias and learns more balanced performance for different races.

show abstract

Section: Debiased Algorithmsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Mitigate Bias in Face Recognition using Skewness-Aware Reinforcement Learning

Wang¹,

Deng²

2019

Preprint

View full text Add to dashboard Cite

show abstract

“…This realworld gender distribution skew becomes part of the data that trains models to recognize or reason about these activities. 1 Naturally, these models then learn discriminative cues which include the gender of the actors. In fact, the gender correlation may even become amplified in the model, as Zhao et al [46] demonstrates.…”

Section: Introductionmentioning

confidence: 99%

“…In this work, we set out to provide an in-depth look at this problem of training visual classifiers in the presence of spurious correlations. We are inspired by prior work on machine learning fairness [45,46,36,1] and aim to build a unified understanding of the proposed techniques. Code is available at https://github.com/ princetonvisualai/DomainBiasMitigation.…”

Section: Introductionmentioning

confidence: 99%

“…Using this benchmark, we demonstrate that the presence of spurious bias in the training data severely degrades the 1 Buolamwini and Gebru [6] note that collecting a more representative training dataset should be the first step of the solution. That is true in the cases they consider (where people with darker skin tones are dramatically and unreasonably undersampled in datasets) but may not be a viable approach to cases where the datasets accurately reflect the real-world skew.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Towards Fairness in Visual Recognition: Effective Strategies for Bias Mitigation

Wang

Qinami

Karakozis

et al. 2019

Preprint

View full text Add to dashboard Cite

Computer vision models learn to perform a task by capturing relevant statistics from training data. It has been shown that models learn spurious age, gender, and race correlations when trained for seemingly unrelated tasks like activity recognition or image captioning. Various mitigation techniques have been presented to prevent models from utilizing or learning such biases. However, there has been little systematic comparison between these techniques. We design a simple but surprisingly effective visual recognition benchmark for studying bias mitigation. Using this benchmark, we provide a thorough analysis of a wide range of techniques. We highlight the shortcomings of popular adversarial training approaches for bias mitigation, propose a simple but similarly effective alternative to the inference-time Reducing Bias Amplification method of Zhao et al., and design a domain-independent training technique that outperforms all other methods. Finally, we validate our findings on the attribute classification task in the CelebA dataset, where attribute presence is known to be correlated with the gender of people in the image, and demonstrate that the proposed technique is effective at mitigating real-world gender bias.

show abstract

Learning Not to Learn: Training Deep Neural Networks With Biased Data

Kim

et al. 2019

2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

273

307

View full text Add to dashboard Cite

We propose a novel regularization algorithm to train deep neural networks, in which data at training time is severely biased. Since a neural network efficiently learns data distribution, a network is likely to learn the bias information to categorize input data. It leads to poor performance at test time, if the bias is, in fact, irrelevant to the categorization. In this paper, we formulate a regularization loss based on mutual information between feature embedding and bias. Based on the idea of minimizing this mutual information, we propose an iterative algorithm to unlearn the bias information. We employ an additional network to predict the bias distribution and train the network adversarially against the feature embedding network. At the end of learning, the bias prediction network is not able to predict the bias not because it is poorly trained, but because the feature embedding network successfully unlearns the bias information. We also demonstrate quantitative and qualitative experimental results which show that our algorithm effectively removes the bias information from feature embedding.

show abstract

Turning a Blind Eye: Explicit Removal of Biases and Variation from Deep Neural Network Embeddings

Cited by 3 publications

References 0 publications

Mitigate Bias in Face Recognition using Skewness-Aware Reinforcement Learning

Mitigate Bias in Face Recognition using Skewness-Aware Reinforcement Learning

Towards Fairness in Visual Recognition: Effective Strategies for Bias Mitigation

Learning Not to Learn: Training Deep Neural Networks With Biased Data

Contact Info

Product

Resources

About