Simulation link: https://malariadetection.streamlit.app/
To test the model download cell images.
This project aims to classify the maleria desiese based on the cell image. There are two types of cell image used here
one is Parasite and another is non parasite ie. Uninfected. The model is trained using the
Deep Convolutional nueral network with other layers. Model has achieved 97% of accuracy on test dataset
for both type of label ie. Parasite and Uninfected.
________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
conv2d_20 (Conv2D) (None, 48, 48, 64) 4864
conv2d_21 (Conv2D) (None, 48, 48, 64) 102464
batch_normalization_6 (Batc (None, 48, 48, 64) 256
hNormalization)
max_pooling2d_6 (MaxPooling (None, 24, 24, 64) 0
2D)
dropout_6 (Dropout) (None, 24, 24, 64) 0
conv2d_22 (Conv2D) (None, 24, 24, 128) 73856
conv2d_23 (Conv2D) (None, 24, 24, 128) 147584
conv2d_24 (Conv2D) (None, 24, 24, 128) 147584
batch_normalization_7 (Batc (None, 24, 24, 128) 512
hNormalization)
max_pooling2d_7 (MaxPooling (None, 12, 12, 128) 0
2D)
dropout_7 (Dropout) (None, 12, 12, 128) 0
conv2d_25 (Conv2D) (None, 12, 12, 256) 295168
conv2d_26 (Conv2D) (None, 12, 12, 256) 590080
conv2d_27 (Conv2D) (None, 12, 12, 256) 590080
conv2d_28 (Conv2D) (None, 12, 12, 256) 590080
conv2d_29 (Conv2D) (None, 12, 12, 256) 590080
batch_normalization_8 (Batc (None, 12, 12, 256) 1024
hNormalization)
max_pooling2d_8 (MaxPooling (None, 6, 6, 256) 0
2D)
flatten_2 (Flatten) (None, 9216) 0
dropout_8 (Dropout) (None, 9216) 0
dense_4 (Dense) (None, 512) 4719104
dense_5 (Dense) (None, 1) 513
=================================================================
Total params: 7,853,249
Trainable params: 7,852,353
Non-trainable params: 896
_________________________________________________________________
Thank you