Focal Loss for Imbalanced Detection Quiz

Answers and explanations appear as you take the quiz.

1. What core problem in one-stage object detection was focal loss designed to address?
   • Extreme imbalance between rare foreground objects and abundant easy background
   • Slow inference speed of two-stage detectors
   • Lack of labeled training data
   • Overfitting to large objects
2. What is the modulating factor that focal loss multiplies onto cross-entropy?
   • log(p_t)
   • (1 - p_t)^gamma
   • p_t^2 only
   • 1 / p_t
3. What happens to focal loss when the focusing parameter gamma is set to 0?
   • It becomes mean squared error
   • It ignores all positive examples
   • It becomes ordinary cross-entropy
   • It diverges to infinity
4. What does increasing gamma do to easy, well-classified examples?
   • Increases their loss contribution
   • Leaves their contribution unchanged
   • Removes them from the dataset
   • Suppresses their loss contribution more strongly
5. Which detector architecture introduced focal loss?
   • RetinaNet
   • YOLOv1
   • Faster R-CNN
   • AlexNet
6. What role does the alpha parameter typically play alongside gamma?
   • It sets the learning rate
   • It balances the relative weight of positive versus negative classes
   • It controls image resolution
   • It defines the number of anchor boxes
7. For a hard example with p_t close to zero, what is the focal modulating factor (1 − p_t)^gamma likely to do?
   • Drive the factor to zero
   • Turn the example into a positive label
   • Keep the factor relatively close to one so the example retains substantial loss weight
   • Replace the cross-entropy term with alpha
8. A dense detector has many easy background locations. What effect does focal loss aim to have on their combined training influence?
   • Increase their count by duplicating them
   • Remove all positive object examples
   • Make every example receive an identical loss
   • Reduce their dominance so training can focus more on sparse hard examples