Image dataset for fine grain classification monkey species

Create CNN Model and Optimize it using Keras Tuner mkdir ~ kaggle cp kaggle json ~ kaggle chmod 600 ~ kagglekaggle json kaggle datasets download d slothkong10 monkey speci. Create CNN Model and Optimize it using Keras Tuner mkdir ~ kaggle cp kaggle json ~ kaggle chmod 600 ~ kagglekaggle json kaggle datasets download d slothkong10 monkey speci.

## Create CNN Model and Optimize it using Keras Tuner

####!mkdir ~/.kaggle

####!cp kaggle.json ~/.kaggle/

####!chmod 600 ~/.kaggle/kaggle.json

####!kaggle datasets download -d slothkong/10-monkey-species

####! unzip 10-monkey-species.zip

from google colab import files

from google.colab import files

from google.colab import drive

drive.mount('/content/drive',)

Drive already mounted at /content/drive; to attempt to forcibly remount, call drive.mount("/conte

### Optimizing the Keras-Tuner

###!pip install keras-tuner

import tensorflow as tf

from tensorflow import keras

import numpy as np

print(tf. version )

2.5.0

import pandas as pd

import numpy as np

import itertools

from sklearn import metrics

from sklearn.metrics import confusion_matrix

from tensorflow.keras.preprocessing.image import ImageDataGenerator, img_to_array, load_img

from tensorflow.keras.models import Sequential

from tensorflow.keras.layers import Conv2D,Flatten,Dropout,Dense

from tensorflow.keras.optimizers import Adam

from tensorflow import keras

from tensorflow.keras import layers

from kerastuner.tuners import RandomSearch

from tensorflow.compat.v1 import ConfigProto

from tensorflow.compat.v1 import InteractiveSession

config = ConfigProto()

config.gpu_options.per_process_gpu_memory_fraction = 0.5

config.gpu_options.allow_growth = True

session = InteractiveSession(config=config)

/usr/local/lib/python3.7/dist-packages/tensorflow/python/client/session.py:1761: UserWarning: An warnings.warn('An interactive session is already active This can '

# Convolutional Neural Network

# Importing the libraries

import tensorflow as tf

from tensorflow.keras.preprocessing.image import ImageDataGenerator

#Default dimensions we found online

img_width, img_height = 224, 224

#Create a bottleneck file

top_model_weights_path = 'optimize_fc_model.h5'

# loading up our datasets

train_data_dir = '/content/training/training'

validation_data_dir = '/content/validation/validation'

# test_data_dir = '/content/test'

# number of epochs to train top model

epochs = 7 #this has been changed after multiple model run

# batch size used by flow_from_directory and predict_generator batch_size = 50

# loading training data

train_datagen = ImageDataGenerator(

rescale=1./255,

shear_range=0.2,

zoom_range=0.2,

horizontal_flip=True)

train_generator = train_datagen.flow_from_directory(

'/content/training/training',

target_size=(64, 64),

batch_size=32,

class_mode='categorical')

Found 1098 images belonging to 11 classes

# loading testing data

test_datagen = ImageDataGenerator(rescale=1./255)

test_generator = train_datagen.flow_from_directory(

'/content/validation/validation',

target_size=(64, 64),

batch_size=32,

class_mode='categorical')

Found 272 images belonging to 11 classes

# initialising sequential model and adding layers to it

cnn = tf.keras.models.Sequential()

cnn.add(tf.keras.layers.Conv2D(filters=48, kernel_size=3, activation='relu', input_shape=[64, 64, 3])) cnn.add(tf.keras.layers.MaxPool2D(pool_size=2, strides=2))

cnn.add(tf.keras.layers.Conv2D(filters=48, kernel_size=3, activation='relu'))

cnn.add(tf.keras.layers.MaxPool2D(pool_size=2, strides=2))

cnn.add(tf.keras.layers.Conv2D(filters=32, kernel_size=3, activation='relu'))

cnn.add(tf.keras.layers.MaxPool2D(pool_size=2, strides=2))

cnn.add(tf.keras.layers.Flatten())

cnn.add(tf.keras.layers.Dense(128, activation='relu'))

cnn.add(tf.keras.layers.Dense(64, activation='relu'))

cnn.add(tf.keras.layers.Dense(11, activation='softmax'))

cnn.compile(optimizer="adam", loss="categorical_crossentropy", metrics=["accuracy"])

history = cnn fit(x=train generator validation data=test generator epochs=25 batch size=50)

history = cnn.fit(x=train_generator, validation_data=test_generator,epochs=25,batch_size=50)

Epoch 1/25

35/35 [==============================] - 36s 1s/step - loss: 2.3320 - accuracy: 0.1485 - val_loss Epoch 2/25

35/35 [==============================] - 34s 998ms/step - loss: 2.0866 - accuracy: 0.2468 - val_l Epoch 3/25

35/35 [==============================] - 34s 957ms/step - loss: 1.8086 - accuracy: 0.3679 - val_l Epoch 4/25

35/35 [==============================] - 34s 973ms/step - loss: 1.6248 - accuracy: 0.4308 - val_l Epoch 5/25

35/35 [==============================] - 35s 995ms/step - loss: 1.4666 - accuracy: 0.4836 - val_l Epoch 6/25

35/35 [==============================] - 35s 1s/step - loss: 1.3259 - accuracy: 0.5410 - val_loss Epoch 7/25

35/35 [==============================] - 35s 985ms/step - loss: 1.2201 - accuracy: 0.5893 - val_l Epoch 8/25

35/35 [==============================] - 36s 1s/step - loss: 1.2016 - accuracy: 0.5783 - val_loss Epoch 9/25

35/35 [==============================] - 35s 1s/step - loss: 1.1860 - accuracy: 0.5801 - val_loss Epoch 10/25

35/35 [==============================] - 35s 1s/step - loss: 1.0415 - accuracy: 0.6403 - val_loss Epoch 11/25

35/35 [==============================] - 35s 991ms/step - loss: 0.9906 - accuracy: 0.6494 - val_l Epoch 12/25

35/35 [==============================] - 35s 983ms/step - loss: 1.0043 - accuracy: 0.6384 - val_l Epoch 13/25

35/35 [==============================] - 34s 990ms/step - loss: 0.8997 - accuracy: 0.6885 - val_l Epoch 14/25

35/35 [==============================] - 35s 1s/step - loss: 0.8433 - accuracy: 0.7022 - val_loss Epoch 15/25

35/35 [==============================] - 36s 1s/step - loss: 0.8521 - accuracy: 0.7067 - val_loss Epoch 16/25

35/35 [==============================] - 37s 1s/step - loss: 0.7924 - accuracy: 0.7177 - val_loss Epoch 17/25

35/35 [==============================] - 36s 1s/step - loss: 0.7527 - accuracy: 0.7432 - val_loss Epoch 18/25

35/35 [==============================] - 37s 1s/step - loss: 0.7653 - accuracy: 0.7195 - val_loss Epoch 19/25

35/35 [==============================] - 37s 1s/step - loss: 0.6716 - accuracy: 0.7687 - val_loss Epoch 20/25

35/35 [==============================] - 36s 1s/step - loss: 0.7026 - accuracy: 0.7623 - val_loss Epoch 21/25

35/35 [==============================] - 36s 1s/step - loss: 0.6732 - accuracy: 0.7659 - val_loss Epoch 22/25

35/35 [==============================] - 35s 1s/step - loss: 0.6195 - accuracy: 0.7805 - val_loss Epoch 23/25

35/35 [==============================] - 34s 965ms/step - loss: 0.5811 - accuracy: 0.7951 - val_l Epoch 24/25

35/35 [==============================] - 36s 1s/step - loss: 0.5338 - accuracy: 0.8224 - val_loss Epoch 25/25

35/35 [==============================] - 35s 1s/step - loss: 0.4787 - accuracy: 0.8279 - val_loss

# summarize history for accuracy

plt.plot(history.history['accuracy'])

plt.plot(history.history['val_accuracy'])

plt.title('model accuracy')

plt ylabel('accuracy')

import matplotlib.pyplot as plt

plt.ylabel( accuracy )

plt.xlabel('epoch')

plt.legend(['train', 'test'], loc='upper left') plt.show()

# summarize history for loss

plt.plot(history.history['loss'])

plt.plot(history.history['val_loss'])

plt.title('model loss')

plt.ylabel('loss')

plt.xlabel('epoch')

plt.legend(['train', 'test'], loc='upper left')

# Evaluating model on validation data

evaluate = cnn.evaluate(test_generator)

print(evaluate)

9/9 [==============================] - 8s 875ms/step - loss: 0.9517 - accuracy: 0.7096 [0.9517170190811157, 0.7095588445663452]

def give_accuracy():

p=cnn.predict(test_generator)

cm=confusion_matrix(y_true=test_generator.classes,y_pred=np.argmax(p,axis=-1))

acc=cm.trace()/cm.sum()

print('The Classification Report \n', cm)

print(f'Accuracy: {acc*100}')

give_accuracy()

The Classification Report

[[2 2 1 6 1 1 3 4 2 4]

[2 1 2 3 0 6 4 4 4 2]

[3 5 3 1 2 2 3 5 3 0]

[3 4 3 1 3 6 3 3 0 4]

[2 4 1 1 3 4 3 5 1 2]

[2 1 2 4 2 3 1 3 3 7]

[1 1 3 3 4 1 3 6 3 1]

[3 2 5 5 2 1 1 6 1 2]

[4 2 6 3 1 3 2 3 3 0]

[4 2 3 5 1 2 4 2 0 3]]

Accuracy: 10.294117647058822

import numpy as np

from tensorflow.keras.preprocessing import image

test_image = image.load_img('/content/validation/validation/n1/n100.jpg', target_size = (64,64)) test_image = image.img_to_array(test_image)

test_image=test_image/255

test_image = np.expand_dims(test_image, axis = 0)

result = cnn.predict(test_image)

test = np.array(test_image)

# making predictions

#prediction = np.argmax(cnn.predict(test_image), axis=-1)

prediction = np.argmax(cnn.predict(test_image))

prediction

8

output = { 0:'mantled_howler',1:'patas_monkey',2:'bald_uakari',3:'japanese_macaque',4:'pygmy_marmoset',

print("The prediction Of the Image is : ", output[prediction])

The prediction Of the Image is : black_headed_night_monkey

# show the image

import matplotlib.pyplot as plt

test_image = image.load_img('/content/validation/validation/n1/n100.jpg', target_size = (64,64))

plt.axis('off')

plt.imshow(test_image)

plt.show()

 0s completed at 2:14 PM