We present Augmented Grad-CAM, a general framework to provide a high-resolution visual explanation of CNN outputs. Our idea is to take advantage of image augmentation to aggregate multiple low-resolution heat-maps - in our experiments Grad-CAMs - computed from augmented copies of the same input image. We generate the high-resolution heat-map through super-resolution, and we formulate a general optimization problem based on Total Variation regularization. This problem is entirely solved on the GPU at inference time, together with image augmentation. Augmented Grad-CAM outperforms Grad-CAM in weakly supervised localization on Imagenet dataset, and provides more detailed heat-maps. Moreover, Augmented Grad-CAM turns to be particularly useful in monitoring the production of silicon wafers, where CNNs are employed to classify defective patterns on the wafer surface to detect harmful faults in the production line.

Augmented Grad-CAM: Heat-Maps Super Resolution Through Augmentation

Carrera D.;Fragneto P.;Boracchi G.
2020-01-01

Abstract

We present Augmented Grad-CAM, a general framework to provide a high-resolution visual explanation of CNN outputs. Our idea is to take advantage of image augmentation to aggregate multiple low-resolution heat-maps - in our experiments Grad-CAMs - computed from augmented copies of the same input image. We generate the high-resolution heat-map through super-resolution, and we formulate a general optimization problem based on Total Variation regularization. This problem is entirely solved on the GPU at inference time, together with image augmentation. Augmented Grad-CAM outperforms Grad-CAM in weakly supervised localization on Imagenet dataset, and provides more detailed heat-maps. Moreover, Augmented Grad-CAM turns to be particularly useful in monitoring the production of silicon wafers, where CNNs are employed to classify defective patterns on the wafer surface to detect harmful faults in the production line.
2020
ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings
978-1-5090-6631-5
Class Activation Mapping
Image Augmentation
Superresolution
Visual interpretation
File in questo prodotto:
File Dimensione Formato  
2020_Augmented_Grad_CAM_ICASSP.pdf

accesso aperto

: Post-Print (DRAFT o Author’s Accepted Manuscript-AAM)
Dimensione 7.53 MB
Formato Adobe PDF
7.53 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: https://hdl.handle.net/11311/1156462
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 16
  • ???jsp.display-item.citation.isi??? 12
social impact