Diagnosis of Cervical Cancer based on Ensemble Deep Learning Network using Colposcopy Images

Table 2

Description of network architecture of the CYENET model.

Layer No.	Layer type	Filter size	Stride	No. of filters	FC units	Input	Output
1	Convolution 1			64	—
2	Max-pool_1			—	—
3	Convolution 2			64	—
4	Convolution 3			128	—
5	Max-pool_2			—	—
6	Parallel convolution 1	, ,		32 ⊕ 64 ⊕ 128	—
7	Max-pool_3			—	—
8	Parallel convolution 2	, ,		32 ⊕ 64 ⊕ 128	—
9	Parallel convolution 3	, ,		32 ⊕ 64 ⊕ 128	—
10	Max-pool_4			—	—
11	Parallel convolution 4	, ,		32 ⊕ 64 ⊕ 128	—
12	Max-pool_5			—	—
13	Fully connected 1	—	—	—	512
14	Fully connected 2	—	—	—	3