Artificial Intelligence Computer Vision 1 — Questions and Answers
Question 1: What is the primary purpose of a convolutional neural network (CNN) in computer vision tasks?
- To translate images into text
- To automatically learn hierarchical spatial features from images for recognition tasks (Correct answer)
- To compress images for storage
- To index images in a database
Correct answer: To automatically learn hierarchical spatial features from images for recognition tasks
CNNs use learnable convolutional filters to extract local spatial features at multiple scales, building hierarchical representations from edges to complex objects.
Question 2: What does 'image segmentation' mean in computer vision?
- Dividing an image into a grid of equal tiles
- Assigning a class label to every pixel in the image (Correct answer)
- Detecting and drawing bounding boxes around objects
- Resizing an image to a standard resolution
Correct answer: Assigning a class label to every pixel in the image
Image segmentation partitions an image into regions, assigning class labels to individual pixels for semantic segmentation or instance boundaries for instance segmentation.
Question 3: Which computer vision task outputs both a class label and a bounding box for each detected object?
- Image classification
- Semantic segmentation
- Object detection (Correct answer)
- Optical flow estimation
Correct answer: Object detection
Object detection identifies what objects are present and where they are by predicting class labels and bounding box coordinates for each detected instance.
Question 4: What does data augmentation in computer vision typically involve?
- Collecting additional labeled images from the internet
- Applying random transformations like flips, rotations, and crops to training images to improve generalization (Correct answer)
- Compressing training images to reduce disk usage
- Converting images to grayscale before training
Correct answer: Applying random transformations like flips, rotations, and crops to training images to improve generalization
Data augmentation artificially expands training data by applying label-preserving transformations, helping models become robust to variations in orientation, scale, and lighting.
Question 5: What is the role of anchor boxes in object detection models like YOLO and Faster R-CNN?
- They define the backbone CNN architecture
- They serve as pre-defined reference bounding boxes of various aspect ratios that the model adjusts to fit detected objects (Correct answer)
- They store the training loss for each image
- They control the color normalization of input images
Correct answer: They serve as pre-defined reference bounding boxes of various aspect ratios that the model adjusts to fit detected objects
Anchor boxes are a set of predefined bounding box shapes; the detector predicts offsets from these anchors to localize objects of different sizes and aspect ratios.
Question 6: What is 'image classification' in computer vision?
- Segmenting an image into foreground and background
- Assigning a single label to an entire image based on its content (Correct answer)
- Detecting and tracking multiple objects across video frames
- Estimating the 3D pose of objects in a scene
Correct answer: Assigning a single label to an entire image based on its content
Image classification takes an image as input and outputs a single class label representing the dominant subject or scene category.
Artificial Intelligence Computer Vision 1
What is the primary purpose of a convolutional neural network (CNN) in computer vision tasks?