salehahmedshafin/pytorch-custom-cnn-model

The architecture is a deep neural network for image processing, composed of convolutional layers capturing features, followed by pooling and fully connected layers. It uses ReLU activation, dropout for regularization, and ends with a linear output layer for classification.

Pytorch-Custom-CNN-model

Dataset-
It is a medical images directory structure branched into 3 subfolders (COVID, NORMAL, PNEUMONIA) which contains the Chest X-ray (CXR) Images.
COVID: 1626 images
NORMAL: 1802 images
PNEUMONIA: 1800 images
Original Dataset Link- https://www.kaggle.com/datasets/sachinkumar413/covid-pneumonia-normal-chest-xray-images
In the Dataset folder contains customized dataset which is splitting the whole dataset into three parts-

1- Training (70%)
2- Validation (15%)
3- Testing (15%)

Model Architecture:

CNN(
(conv_layers): Sequential(
(0): Conv2d(3, 16, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(1): ReLU()
(2): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)
(3): Conv2d(16, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(4): ReLU()
(5): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)
(6): Conv2d(32, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(7): ReLU()
(8): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)
(9): Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(10): ReLU()
(11): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)
)
(fc_layers): Sequential(
(0): Linear(in_features=25088, out_features=512, bias=True)
(1): ReLU()
(2): Dropout(p=0.2, inplace=False)
(3): Linear(in_features=512, out_features=256, bias=True)
(4): ReLU()
(5): Dropout(p=0.2, inplace=False)
(6): Linear(in_features=256, out_features=128, bias=True)
(7): ReLU()
(8): Dropout(p=0.2, inplace=False)
(9): Linear(in_features=128, out_features=3, bias=True)
)
)

There are 4 Conv2d layers and 4 Linear Layer and use MaxPool in the pooling layer

Model Summary:

Layer (type)	Output Shape	Param #
Conv2d-1	[-1, 16, 224, 224]	448
ReLU-2	[-1, 16, 224, 224]	0
MaxPool2d-3	[-1, 16, 112, 112]	0
Conv2d-4	[-1, 32, 112, 112]	4,640
ReLU-5	[-1, 32, 112, 112]	0
MaxPool2d-6	[-1, 32, 56, 56]	0
Conv2d-7	[-1, 64, 56, 56]	18,496
ReLU-8	[-1, 64, 56, 56]	0
MaxPool2d-9	[-1, 64, 28, 28]	0
Conv2d-10	[-1, 128, 28, 28]	73,856
ReLU-11	[-1, 128, 28, 28]	0
MaxPool2d-12	[-1, 128, 14, 14]	0
Linear-13	[-1, 512]	12,845,568
ReLU-14	[-1, 512]	0
Dropout-15	[-1, 512]	0
Linear-16	[-1, 256]	131,328
ReLU-17	[-1, 256]	0
Dropout-18	[-1, 256]	0
Linear-19	[-1, 128]	32,896
ReLU-20	[-1, 128]	0
Dropout-21	[-1, 128]	0
Linear-22	[-1, 3]	387
Total params		13,107,619
Trainable params		13,107,619
Non-trainable params		0
Input size (MB)		0.57
Forward/backward pass size (MB)		25.86
Params size (MB)		50.00
Estimated Total Size (MB)		76.44