Metadata Normalization

@article{Lu2021MetadataN,
  title={Metadata Normalization},
  author={Mandy Lu and Qingyu Zhao and Jiequan Zhang and Kilian M. Pohl and Li Fei-Fei and Juan Carlos Niebles and Ehsan Adeli},
  journal={2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
  year={2021},
  pages={10912-10922}
}
  • Mandy Lu, Qingyu Zhao, E. Adeli
  • Published 19 April 2021
  • Computer Science
  • 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
Batch Normalization (BN) and its variants have delivered tremendous success in combating the covariate shift induced by the training step of deep learning methods. While these techniques normalize the feature distribution by standardizing with batch statistics, they do not correct the influence on features from extraneous variables or multiple distributions. Such extra variables, referred to as metadata here, may create bias or confounding effects (e.g., race when classifying gender from face… 

Figures and Tables from this paper

Revisiting the "Video" in Video-Language Understanding
TLDR
The atemporal probe is proposed, a new model for video-language analysis which provides a stronger bound on the baseline accuracy of multimodal models constrained by image-level understanding, and effectively integrating ATP into full video-level temporal models can improve efficiency and state-of-the-art accuracy.
RELATION INFERENCE AMONG SENSOR TIME SERIES WITH NEURAL NETWORKS IN THE FREQUENCY DOMAIN
TLDR
This project used convolutional neural network and deep metric learning, along with K-cut algorithm, to do spatially grouping among sensors to make the system robust for delay, and the input time-series is converted to the frequencydomain.

References

SHOWING 1-10 OF 73 REFERENCES
Representation Learning with Statistical Independence to Mitigate Bias
  • E. Adeli, Qingyu Zhao, K. Pohl
  • Computer Science
    2021 IEEE Winter Conference on Applications of Computer Vision (WACV)
  • 2021
TLDR
A model based on adversarial training with two competing objectives to learn features that have maximum discriminative power with respect to the task and minimal statistical mean dependence with the protected (bias) variable(s) is proposed.
Group Normalization
TLDR
Group Normalization can outperform its BN-based counterparts for object detection and segmentation in COCO, and for video classification in Kinetics, showing that GN can effectively replace the powerful BN in a variety of tasks.
Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
TLDR
Applied to a state-of-the-art image classification model, Batch Normalization achieves the same accuracy with 14 times fewer training steps, and beats the original model by a significant margin.
Turning a Blind Eye: Explicit Removal of Biases and Variation from Deep Neural Network Embeddings
TLDR
It is demonstrated on this dataset, for a number of facial attribute classification tasks, that the algorithm can be used to remove racial biases from the network feature representation.
Training confounder-free deep learning models for medical applications
TLDR
This article introduces an end-to-end approach for deriving features invariant to confounding factors while accounting for intrinsic correlations between the confounder(s) and prediction outcome, exploiting concepts from traditional statistical methods and recent fair machine learning schemes.
Filter Response Normalization Layer: Eliminating Batch Dependence in the Training of Deep Neural Networks
TLDR
The Filter Response Normalization (FRN) layer is proposed, a novel combination of a normalization and an activation function that can be used as a replacement for other normalizations and activations, and outperforms BN and other alternatives in a variety of settings for all batch sizes.
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.
The Variational Fair Autoencoder
TLDR
This model is based on a variational autoencoding architecture with priors that encourage independence between sensitive and latent factors of variation that is more effective than previous work in removing unwanted sources of variation while maintaining informative latent representations.
Confounder-Aware Visualization of ConvNets
TLDR
An approach that aims to visualize confounder-free saliency maps that only highlight voxels predictive of the diagnosis is proposed, which incorporates univariate statistical tests to identify confounding effects within the intermediate features learned by ConvNet.
Domain-Adversarial Training of Neural Networks
TLDR
A new representation learning approach for domain adaptation, in which data at training and test time come from similar but different distributions, which can be achieved in almost any feed-forward model by augmenting it with few standard layers and a new gradient reversal layer.
...
...