Introduction
This repository is about the image classification using deep learning. To implement the image classification I have worked on the CIFAR 10 dataset which contains 10 class, toatal 60000 images of size 32 x 32 . Download the dataset from here.
Model Architecture
Layer 1:
- Convolution with 32 different filters in size of (3x3)
- Max Pooling by 2
- ReLU activation function
- Batch Normalization
Layer 2:
- Convolution with 64 different filters in size of (3x3)
- Max Pooling by 2
- ReLU activation function
- Batch Normalization
Layer 3:
- Convolution with 128 different filters in size of (3x3)
- Max Pooling by 2
- ReLU activation function
- Batch Normalization
Layer 4:
- Convolution with 256 different filters in size of (3x3)
- Max Pooling by 2
- ReLU activation function
- Batch Normalization
Layer 5:
- Convolution with 512 different filters in size of (3x3)
- Max Pooling by 2
- ReLU activation function
- Batch Normalization
Flattening the 3-D output of the last convolving operations.
Dense Layer with 1024 units
- Dropout(0.2)
- Batch Normalization
Dense Layer with 512 units
- Dropout(0.3)
- Batch Normalization
Dense Layer with 256 units
- Dropout (0.4)
- Batch Normalization
Dense Layer with 128 units
- Dropout (0.5)
- Batch Normalization
Dense Layer with 10 units (number of image classes)
Optimizer : opt_rms
Loss : categorical_crossentropy