Artificial Neural Network (ann) Codes

Group B

QB5: 5. Write Python program to implement CNN object detection. Discuss numerous performance evaluation metrics for evaluating the object detecting algorithms' performance

CNN Object Detection

Solution and implementation for QB5 from Artificial Neural Network (ann).

B5_CNN_obj_detect.ipynb Download

Dog vs Cat Image Classification

This notebook demonstrates a simple binary classification model using TensorFlow and Keras.

Code Cell [In]
!pip install tensorflow opencv-python matplotlib -q

1. Environment Setup and Imports

Installing and importing necessary libraries for computer vision and deep learning.

Code Cell [In]
import cv2
import numpy as np
import matplotlib.pyplot as plt
import tensorflow as tf
from tensorflow.keras import layers, models

2. Dataset Loading

Loading the 'cats_vs_dogs' dataset from TensorFlow Datasets.

Code Cell [In]
import tensorflow_datasets as tfds

dataset, info = tfds.load('cats_vs_dogs', with_info=True, as_supervised=True)

train_data = dataset['train']
Code Cell [In]
plt.figure(figsize=(10,5))

for i, (image, label) in enumerate(train_data.take(5)):
    plt.subplot(1,5,i+1)
    plt.imshow(image)
    plt.title("Dog" if label.numpy() == 1 else "Cat")
    plt.axis("off")

plt.show()

3. Data Preprocessing

Normalizing and resizing images to prepare them for the neural network.

Code Cell [In]
IMG_SIZE = 128

def preprocess(image, label):
    image = tf.image.resize(image, (IMG_SIZE, IMG_SIZE))
    image = image / 255.0
    return image, label

train_data = train_data.map(preprocess).batch(32).take(200)  # small subset

4. Model Architecture

Building a Convolutional Neural Network (CNN) for binary classification.

Code Cell [In]
model = models.Sequential([
    layers.Conv2D(16, (3,3), activation='relu', input_shape=(128,128,3)),
    layers.MaxPooling2D(),

    layers.Conv2D(32, (3,3), activation='relu'),
    layers.MaxPooling2D(),

    layers.Flatten(),
    layers.Dense(64, activation='relu'),
    layers.Dense(1, activation='sigmoid')  # binary classification
])

model.compile(optimizer='adam',
              loss='binary_crossentropy',
              metrics=['accuracy'])

model.summary()

5. Model Training

Training the model on a subset of the data.

Code Cell [In]
history = model.fit(train_data, epochs=3)

6. Inference and Visualization

Defining a prediction function and testing it on an image.

Code Cell [In]
def predict_image(img_path):
    img = cv2.imread(img_path)
    img_resized = cv2.resize(img, (128,128))
    img_norm = img_resized / 255.0
    img_input = np.expand_dims(img_norm, axis=0)

    prediction = model.predict(img_input)[0][0]

    label = "Dog" if prediction > 0.5 else "Cat"
    return img, label
Code Cell [In]
img, label = predict_image("image1.jpg")

# Draw a simple bounding box (whole image)
h, w, _ = img.shape
cv2.rectangle(img, (10,10), (w-10, h-10), (0,255,0), 2)

cv2.putText(img, label, (20,40),
            cv2.FONT_HERSHEY_SIMPLEX, 1,
            (0,255,0), 2)

plt.imshow(cv2.cvtColor(img, cv2.COLOR_BGR2RGB))
plt.axis('off')
Code Cell [In]
img, label = predict_image("image2.jpg")

# Draw a simple bounding box (whole image)
h, w, _ = img.shape
cv2.rectangle(img, (10,10), (w-10, h-10), (0,255,0), 2)

cv2.putText(img, label, (20,40),
            cv2.FONT_HERSHEY_SIMPLEX, 1,
            (0,255,0), 2)

plt.imshow(cv2.cvtColor(img, cv2.COLOR_BGR2RGB))
plt.axis('off')
image1.jpg Download
Answer image: image1.jpg
image2.jpg Download
Answer image: image2.jpg

Other Questions in Artificial Neural Network

See All Available Questions 


        

        

            Download