CNN CNN - Convolutional Neural Networks Data Augmentation Techniques Questions and Answers 1 — Questions and Answers
Question 1: What is the primary purpose of data augmentation in CNN training?
- To increase model size
- To artificially expand the training dataset and improve generalization (Correct answer)
- To speed up inference
- To reduce the number of layers
Correct answer: To artificially expand the training dataset and improve generalization
Data augmentation artificially expands the training dataset by applying transformations, helping CNNs generalize better to unseen data.
Question 2: Which augmentation technique randomly flips an image along its vertical axis?
- Random rotation
- Horizontal flip (Correct answer)
- Vertical flip
- Random crop
Correct answer: Horizontal flip
A horizontal flip mirrors an image left-to-right along the vertical axis, creating a natural augmentation for many image classification tasks.
Question 3: What does the 'random crop' augmentation technique do during CNN training?
- Removes random pixels from the image
- Crops a random sub-region of the image for training (Correct answer)
- Reduces image resolution randomly
- Masks random channels
Correct answer: Crops a random sub-region of the image for training
Random cropping extracts a randomly positioned sub-region of the input image, forcing the CNN to learn from different spatial locations.
Question 4: What is Mixup augmentation in CNN training?
- Mixing different optimizers
- Creating synthetic samples by linearly interpolating between two training images and their labels (Correct answer)
- Randomly mixing color channels
- Combining multiple loss functions
Correct answer: Creating synthetic samples by linearly interpolating between two training images and their labels
Mixup blends two training images and their corresponding labels by a random ratio λ, encouraging the model to learn smoother decision boundaries.
Question 5: What does the CutOut augmentation technique do?
- Removes the image background
- Randomly masks out a square region of the input during training (Correct answer)
- Crops the image into multiple patches
- Applies random color cutoffs
Correct answer: Randomly masks out a square region of the input during training
CutOut randomly zeroes out a square patch in the training image, forcing the network to use context from the entire image rather than relying on a single discriminative region.
Question 6: Which PyTorch library is most commonly used for applying data augmentation transforms to image datasets?
- torch.nn
- torchvision.transforms (Correct answer)
- torch.optim
- torch.utils.data
Correct answer: torchvision.transforms
torchvision.transforms provides a comprehensive set of image augmentation operations such as RandomHorizontalFlip, RandomCrop, and ColorJitter.
CNN CNN - Convolutional Neural Networks Data Augmentation Techniques Questions and Answers 1
What is the primary purpose of data augmentation in CNN training?