Optic Disc and Optic Cup Segmentation for Glaucoma Detection from Blur Retinal Images Using Improved Mask-RCNN

Table 4

Comparative analysis of the proposed approach with base models.

Parameters	Inception-V4	VGG-16	ResNet-101	ResNet-152	DenseNet-121	DenseNet-77
Total parameters (million)	41.2	119.6	42.5	58.5	7.1	6.2
Training loss	0.0102	0.5069	4.1611e^−04	2.4844e^−04	5.6427e^−04	6.442e^−04
Test loss	0.0686	0.6055	0.02082	0.0246	0.0159	0.0085
Training accuracy	99.74%	83.86%	99.99%	100%	100%	100%
Test accuracy	98.08%	81.83%	99.66%	99.59%	99.75%	99.983%
Processing time (s)	4042	1051	2766	4366	2165	1067