Week 2: Object Detection & Recognition Flashcards
What is image pattern recognition?
Image pattern recognition is the process of analyzing visual content to assign meaningful labels.
What is image classification?
Image classification is assigning a single label to an entire image.
What are the advantages of CNNs compared to feature-based classification?
CNNs can learn image features automatically, eliminating the need for hand-crafted features.
What are the three main types of layers in a CNN?
Convolutional layers, pooling layers, and fully-connected layers.
What is the purpose of a convolutional layer?
A convolutional layer applies filters to extract features from images.
What is the purpose of a pooling layer?
A pooling layer reduces the spatial resolution of feature maps.
What is the purpose of a fully-connected layer?
A fully-connected layer performs linear classification based on learned features.
What is the loss function?
The loss function quantifies how close model predictions are to ground truth labels.
What is backpropagation?
Backpropagation is an algorithm used to update model weights based on the loss function.
What is transfer learning?
Transfer learning allows us to reuse weights from a pre-trained network.
What is the difference between object detection, instance segmentation, and semantic segmentation?
Object detection finds objects in an image and draws bounding boxes around them. Instance segmentation segments each individual object and assigns a unique label to each. Semantic segmentation assigns a class label to each pixel in an image.
What are the three ways to evaluate classification performance?
Confusion matrix, accuracy, precision, and recall.
What does CNN stand for in the context of image classification?
CNN stands for Convolutional Neural Network, a specialized neural network used primarily for image-related tasks in machine learning.
What distinguishes classification, detection, semantic segmentation, and instance segmentation in computer vision?
Classification involves categorizing an entire image. Detection identifies and localizes objects within an image. Semantic segmentation assigns class labels to every pixel. Instance segmentation identifies individual instances of objects and segments them separately.
What is the significance of evaluation metrics in various computer vision tasks?
It refers to the set of metrics used to measure the effectiveness of a machine-learning model. These metrics help to determine how well a model is able to make accurate predictions or classifications on unseen data.
How can one utilize a pre-trained Mask R-CNN model?
To use a pre-trained Mask R-CNN model, one can download it from a repository or framework, then load it into their environment and apply it to perform instance segmentation on images or videos.
What are four key differences between classification, segmentation, and object detection tasks?
- Classification deals with categorizing entire images, whereas segmentation involves pixel-level labeling.
- Object detection localizes and identifies objects within an image, while classification predicts categories.
- Segmentation distinguishes individual pixels by class, while detection provides bounding boxes around objects.
- Instance segmentation identifies and delineates individual instances of objects within an image, unlike semantic segmentation that groups pixels by class.
