Artificial neural networks perform state-of-the-art in an ever-growing number of tasks, and nowadays they are used to solve an incredibly large variety of tasks. There are problems, like the presence of biases in the training data, which question the generalization capability of these models. In this work we propose EnD, a regularization strategy whose aim is to prevent deep models from learning unwanted biases. In particular, we insert an “information bottleneck” at a certain point of the deep neural network, where we disentangle the information about the bias, still letting the useful information for the training task forward-propagating in the rest of the model. One big advantage of EnD is that it does not require additional training complexity (like decoders or extra layers in the model), since it is a regularizer directly applied on the trained model. Our experiments show that EnD 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.

EnD: Entangling and Disentangling Deep Representations for Bias Correction

Enzo Tartaglione;Carlo Alberto Barbano;Marco Grangetto
2021

Abstract

Artificial neural networks perform state-of-the-art in an ever-growing number of tasks, and nowadays they are used to solve an incredibly large variety of tasks. There are problems, like the presence of biases in the training data, which question the generalization capability of these models. In this work we propose EnD, a regularization strategy whose aim is to prevent deep models from learning unwanted biases. In particular, we insert an “information bottleneck” at a certain point of the deep neural network, where we disentangle the information about the bias, still letting the useful information for the training task forward-propagating in the rest of the model. One big advantage of EnD is that it does not require additional training complexity (like decoders or extra layers in the model), since it is a regularizer directly applied on the trained model. Our experiments show that EnD 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.
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
virtual
June 19-25
Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
IEEE/CVF
13508
13517
https://openaccess.thecvf.com/content/CVPR2021/html/Tartaglione_EnD_Entangling_and_Disentangling_Deep_Representations_for_Bias_Correction_CVPR_2021_paper.html
Enzo Tartaglione, Carlo Alberto Barbano, Marco Grangetto
File in questo prodotto:
File Dimensione Formato  
Tartaglione_EnD_Entangling_and_Disentangling_Deep_Representations_for_Bias_Correction_CVPR_2021_paper.pdf

accesso aperto

Tipo di file: PDF EDITORIALE
Dimensione 1.09 MB
Formato Adobe PDF
1.09 MB Adobe PDF Visualizza/Apri

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: http://hdl.handle.net/2318/1795476
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 4
  • ???jsp.display-item.citation.isi??? 2
social impact