• Corpus ID: 239049762

Does Data Repair Lead to Fair Models? Curating Contextually Fair Data To Reduce Model Bias

@article{Agarwal2021DoesDR,
  title={Does Data Repair Lead to Fair Models? Curating Contextually Fair Data To Reduce Model Bias},
  author={Sharat Agarwal and Sumanyu Muku and Saket Anand and Chetan Arora},
  journal={ArXiv},
  year={2021},
  volume={abs/2110.10389}
}
Contextual information is a valuable cue for Deep Neural Networks (DNNs) to learn better representations and improve accuracy. However, co-occurrence bias in the training dataset may hamper a DNN model’s generalizability to unseen scenarios in the real world. For example, in COCO [26], many object categories have a much higher cooccurrence with men compared to women, which can bias a DNN’s prediction in favor of men. Recent works have focused on task-specific training strategies to handle bias… 

Figures and Tables from this paper

References

SHOWING 1-10 OF 58 REFERENCES
REPAIR: Removing Representation Bias by Dataset Resampling
  • Y. Li, N. Vasconcelos
  • Computer Science
    2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
  • 2019
TLDR
Experiments with synthetic and action recognition data show that dataset REPAIR can significantly reduce representation bias, and lead to improved generalization of models trained on REPAired datasets.
Don’t Judge an Object by Its Context: Learning to Overcome Contextual Bias
TLDR
This work focuses on addressing contextual biases to improve the robustness of the learnt feature representations of a category from its co-occurring context by learning a feature subspace that explicitly represents categories occurring in the absence of context along side a joint feature sub space that represents both categories and context.
Contextual Diversity for Active Learning
TLDR
The notion of contextual diversity that captures the confusion associated with spatially co-occurring classes is introduced and state of the art results for active learning on benchmark datasets of Semantic Segmentation, Object Detection and Image Classification are established.
EnD: Entangling and Disentangling deep representations for bias correction
TLDR
EnD, a regularization strategy whose aim is to prevent deep models from learning unwanted biases, is proposed, which effectively improves the generalization on unbiased test sets, and it can be effectively applied on real-case scenarios, like removing hidden biases in the COVID-19 detection from radiographic images.
Towards Fairness in Visual Recognition: Effective Strategies for Bias Mitigation
TLDR
A simple but surprisingly effective visual recognition benchmark for studying bias mitigation, and 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.
Women also Snowboard: Overcoming Bias in Captioning Models
TLDR
A new Equalizer model is introduced that ensures equal gender probability when gender Evidence is occluded in a scene and confident predictions when gender evidence is present and has lower error than prior work when describing images with people and mentioning their gender and more closely matches the ground truth ratio of sentences including women to sentences including men.
Balanced Datasets Are Not Enough: Estimating and Mitigating Gender Bias in Deep Image Representations
TLDR
It is shown that trained models significantly amplify the association of target labels with gender beyond what one would expect from biased datasets, and an adversarial approach is adopted to remove unwanted features corresponding to protected variables from intermediate representations in a deep neural network.
Learning Loss for Active Learning
  • Donggeun Yoo, I. Kweon
  • Computer Science
    2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
  • 2019
TLDR
A novel active learning method that is simple but task-agnostic, and works efficiently with the deep networks, by attaching a small parametric module, named ``loss prediction module,'' to a target network, and learning it to predict target losses of unlabeled inputs.
Can Active Learning Preemptively Mitigate Fairness Issues?
TLDR
This paper study whether models trained with uncertainty-based AL heuristics such as BALD are fairer in their decisions with respect to a protected class than those trained with identically independently distributed (i.i.d.) sampling.
Men Also Like Shopping: Reducing Gender Bias Amplification using Corpus-level Constraints
TLDR
This work proposes to inject corpus-level constraints for calibrating existing structured prediction models and design an algorithm based on Lagrangian relaxation for collective inference to reduce the magnitude of bias amplification in multilabel object classification and visual semantic role labeling.
...
1
2
3
4
5
...