Metadata Normalization

Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)(2021)

引用 9|浏览23
暂无评分
摘要
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 feature distributions 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 images). We introduce the Metadata Normalization (MDN) layer, a new batch-level operation which can be used end-to-end within the training framework, to correct the influence of metadata on feature distributions. MDN adopts a regression analysis technique traditionally used for preprocessing to remove (regress out) the metadata effects on model features during training. We utilize a metric based on distance correlation to quantify the distribution bias from the metadata and demonstrate that our method successfully removes metadata effects on four diverse settings: one synthetic, one 2D image, one video, and one 3D medical image dataset.
更多
查看译文
关键词
batch normalization,BN,covariate shift,deep learning methods,feature distribution,batch statistics,extraneous variables,gender classification,face images,MDN,batch-level operation,training framework,regression analysis technique,metadata effects,model features,distribution bias,metadata normalization layer
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要