ORCNet: A context-based network to simultaneously segment the ocular region components

TL;DR

ORCNet is a novel context-based neural network for precise ocular region segmentation, utilizing a unique loss function that incorporates semantic, spatial, and scale relationships, outperforming existing methods.

Contribution

The paper introduces ORCNet with a new PC-Loss function that leverages semantic relationships for improved ocular segmentation accuracy.

Findings

ORCNet outperforms baseline techniques in F-Score, Error Rate, and IoU.

The method achieves up to 28.26% improvement in Error Rate.

Provides a large dataset of manually labeled masks for research.

Abstract

Accurate extraction of the Region of Interest is critical for successful ocular region-based biometrics. In this direction, we propose a new context-based segmentation approach, entitled Ocular Region Context Network (ORCNet), introducing a specific loss function, i.e., he Punish Context Loss (PC-Loss). The PC-Loss punishes the segmentation losses of a network by using a percentage difference value between the ground truth and the segmented masks. We obtain the percentage difference by taking into account Biederman's semantic relationship concepts, in which we use three contexts (semantic, spatial, and scale) to evaluate the relationships of the objects in an image. Our proposal achieved promising results in the evaluated scenarios: iris, sclera, and ALL (iris + sclera) segmentations, utperforming the literature baseline techniques. The ORCNet with ResNet-152 outperforms the best baseline (EncNet with ResNet-152) on average by 2.27%, 28.26% and 6.43% in terms of F-Score, Error Rate and Intersection Over Union, respectively. We also provide (for research purposes) 3,191 manually labeled masks for the MICHE-I database, as another contribution of our work.