# FPD-M-net: Fingerprint Image Denoising and Inpainting Using M-Net Based Convolutional Neural Networks

## TL;DR

This paper introduces FPD-M-net, a CNN-based architecture that effectively denoises and inpaints degraded fingerprint images by framing the task as segmentation, utilizing a structure similarity loss for improved accuracy.

## Contribution

The paper presents a novel M-net based CNN architecture with a structure similarity loss for fingerprint denoising and inpainting, achieving competitive performance.

## Key findings

- Outperforms baseline methods in fingerprint denoising and inpainting
- Achieved 3rd place in ECCV 2018 Chalearn LAP Competition
- Demonstrates effectiveness of structure similarity loss in segmentation tasks

## Abstract

Fingerprint is a common biometric used for authentication and verification of an individual. These images are degraded when fingers are wet, dirty, dry or wounded and due to the failure of the sensors, etc. The extraction of the fingerprint from a degraded image requires denoising and inpainting. We propose to address these problems with an end-to-end trainable Convolutional Neural Network based architecture called FPD-M-net, by posing the fingerprint denoising and inpainting problem as a segmentation (foreground) task. Our architecture is based on the M-net with a change: structure similarity loss function, used for better extraction of the fingerprint from the noisy background. Our method outperforms the baseline method and achieves an overall 3rd rank in the Chalearn LAP Inpainting Competition Track 3 - Fingerprint Denoising and Inpainting, ECCV 2018