Scribd
Upload
77 views11 pages

AlexNet Model Implementation Guide

The document describes the architecture of AlexNet, an influential convolutional neural network model for image classification. It contains 12 layers including convolutional layers, max pooling layers, and fully connected layers. The model takes 227x227x3 RGB images as input and outputs predictions for 1000 classes through sequential convolutional, pooling and fully connected layers. It is implemented and trained on CIFAR10 dataset using Keras to classify images with over 90% accuracy.

Uploaded by

Mackrish Malik
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
77 views11 pages

AlexNet Model Implementation Guide

The document describes the architecture of AlexNet, an influential convolutional neural network model for image classification. It contains 12 layers including convolutional layers, max pooling layers, and fully connected layers. The model takes 227x227x3 RGB images as input and outputs predictions for 1000 classes through sequential convolutional, pooling and fully connected layers. It is implemented and trained on CIFAR10 dataset using Keras to classify images with over 90% accuracy.

Uploaded by

Mackrish Malik
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 11

Alexnet

DR. DEEPAK MEHTA


Basic Intuition
Input Image Size: 227X227X3
Conv1: 11X11, Stride:4, No. of Filters: 96  55X55X96
Maxpooling: Kernel: 3X3, Stride:2 27X27X96
Conv2: 5X5, Stride:1, No. of Filters:256,padding: 2 27X27X256
Maxpooling: 3X3, Stride:2 13X13X256
Conv3: 3X3, pad:1, no. of filters: 384,padding:1 13X13X384
Conv4: 3X3, pad:1, no. of filters:384, padding:1 13X13X384
Conv5: 3X3,pad:1, no. of fiters: 256 13X13X256
Maxpooling: 3X3, Stride:26X6X256 FCFC
Import Libraries
import numpy as np
import tensorflow as tf
from tensorflow import keras
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Conv2D, MaxPooling2D, Flatten, Dense, BatchNormalization,
Dropout
Model Implementation
model = Sequential()
# Layer 1: Convolutional Layer
model.add(Conv2D(96, (11, 11), strides=(4, 4), input_shape=(227, 227, 3), activation='relu'))
# Layer 2: Max Pooling
model.add(MaxPooling2D(pool_size=(3, 3), strides=(2, 2)))
# Layer 3: Convolutional Layer
model.add(Conv2D(256, (5, 5), padding='same', activation='relu'))
# Layer 4: Max Pooling
model.add(MaxPooling2D(pool_size=(3, 3), strides=(2, 2)))
# Layer 5: Convolutional Layer
model.add(Conv2D(384, (3, 3), padding='same', activation='relu'))
# Layer 6: Convolutional Layer
model.add(Conv2D(384, (3, 3), padding='same', activation='relu'))
# Layer 7: Convolutional Layer
model.add(Conv2D(256, (3, 3), padding='same', activation='relu'))
# Layer 8: Max Pooling
model.add(MaxPooling2D(pool_size=(3, 3), strides=(2, 2))
# Layer 9: Flatten
model.add(Flatten())
# Layer 10: Fully Connected Layer
model.add(Dense(4096, activation='relu'))
model.add(Dropout(0.5))
# Layer 11: Fully Connected Layer
model.add(Dense(4096, activation='relu'))
model.add(Dropout(0.5))
# Layer 12: Output Layer
model.add(Dense(1000, activation='softmax')) # 1000 classes in the ImageNet dataset
# Compile the model
model.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy'])
# Print the model summary
model.summary()
Model Training
(x_train, y_train), (x_test, y_test) = keras.datasets.cifar10.load_data()
x_train, x_test = x_train / 255.0, x_test / 255.0
datagen = ImageDataGenerator(
rotation_range=15,
width_shift_range=0.1,
height_shift_range=0.1,
horizontal_flip=True,
zoom_range=0.2,
)
datagen.fit(x_train)
model.fit(datagen.flow(x_train, y_train, batch_size=32), steps_per_epoch=len(x_train) / 32,
epochs=10, validation_data=(x_test, y_test))
test_loss, test_accuracy = model.evaluate(x_test, y_test, verbose=2)
print(f"Test accuracy: {test_accuracy*100:.2f}%")
print(f"Test loss: {test_loss:.4f}")
Thanks
[email protected]
9812881611

You might also like