NEW
Font size
WorksheetsRecap Object Detection
Total questions: 8
Worksheet time: 4mins
Which of the following is NOT a commonly used object detection algorithm?
YOLO
SSD
R-CNN
ResNet
What does the "anchor box" represent in object detection models like YOLO and SSD?
A ground truth label
A proposed region for classification
A pre-defined box used to predict bounding boxes
A type of data augmentation
What is the purpose of the Intersection over Union (IoU) metric?
To improve model training time
To resize input images correctly
To measure overlap between two boxes
To detect the object class name
Which of the following best describes the YOLO detection approach?
Detects objects in multiple stages
Uses a grid to predict bounding boxes
Uses selective region proposals
Predicts bounding boxes with a decoder
In Faster R-CNN, what does the RPN do?
It classifies object categories
It generates object region proposals
It filters out noisy bounding boxes
It calculates total model loss
What is the function of Non-Maximum Suppression (NMS)?
Combines object feature vectors
Eliminates duplicate box predictions
Reduces training dataset size
Converts masks to bounding boxes
Which metric is most commonly used to evaluate object detection models?
Accuracy over all images
Average classification score
Mean Average Precision (mAP)
Mean Square Bounding Error
What is a major difference between classification and object detection?
Detection classifies without labels
Detection uses deeper neural nets
Detection finds object locations too
Detection avoids feature extraction
