Computer Vision – Image Classification (Day 3 - Test 1)

What is the primary goal of image classification in computer vision?

• Enhance image resolution

• Label an image with a specific category

• Compress the image

• Convert image to grayscale

Which of the following is a popular dataset used for benchmarking image classification models?

• MNIST

• CIFAR-10

• COCO Caption

• UCI Heart Disease

What is the role of a Convolutional Neural Network (CNN) in image classification?

• Generate random features

• Flatten images into single vectors

• Extract hierarchical features from images

• Perform natural language processing

In a CNN, what does the pooling layer typically do?

• Increases image size

• Converts image to binary

• Normalizes pixel values

• Reduces spatial dimensions

What type of activation function is commonly used in image classification tasks within CNNs?

• Sigmoid

• ReLU

• Softmax

• Tanh

What does the Softmax function do in the output layer of an image classification model?

• Converts outputs into probabilities

• Reduces overfitting

• Normalizes input values

• Enhances image contrast

Which loss function is commonly used for multi-class image classification?

• Hinge Loss

• Mean Absolute Error (MAE)

• Binary Cross-Entropy

• Categorical Cross-Entropy

Which technique can help reduce overfitting in image classification models?

• Increasing the learning rate

• Removing dropout layers

• Data augmentation

• Reducing dataset size

What does transfer learning involve in image classification?

• Sharing weights between neurons in the same layer

• Using pre-trained models for new tasks

• Increasing the model depth

• Transforming color spaces

Why is image normalization important before training an image classification model?

• To remove the color from the image

• To increase the size of the dataset

• To ensure all pixel values are on a similar scale

• To convert images into grayscale