AI in Crop Disease Diagnosis Quiz

Answers and explanations appear as you take the quiz.

1. What image-classification approach did Mohanty et al. evaluate on the PlantVillage leaf-image dataset?
   • A convolutional neural network trained to classify crop-disease images
   • A satellite-only weather forecast model
   • A recurrent speech recognition network
   • A hand-written treatment rule with no image input
2. In a multispectral study of gray mold in tomato leaves, why was near-infrared reflectance measured?
   • To measure the temperature of the soil beneath each leaf
   • To test whether spectral changes could be observed before visible symptoms
   • To identify the plant species from its roots
   • To replace all disease labels with weather forecasts
3. What does NDVI indicate when computed from red and near-infrared reflectance?
   • A definitive diagnosis of a particular plant disease
   • The exact amount of pesticide needed
   • A vegetation greenness or condition signal
   • The number of pathogens in a leaf
4. What does domain shift mean for a crop-image model?
   • The model automatically changes the target crop
   • Images receive different file names
   • A disease moves from root to leaf
   • Training and deployment data differ in distribution, which can change model performance
5. What potential benefit can transfer learning offer when labeled crop images are limited?
   • A pretrained model may provide useful features to adapt to the target task
   • It guarantees better results for any crop and disease
   • It removes the need for field validation
   • It makes training data unnecessary
6. Roughly what share of global crop yields is lost to pests and diseases each year?
   • Less than 1%
   • About 20-40%
   • Exactly 5%
   • Over 90%
7. How many controlled-condition leaf images did the PlantVillage paper use?
   • 54,306 images
   • 5,430 images
   • 543 images
   • 540,306 images
8. How is NDVI calculated from near-infrared and red reflectance?
   • (NIR + red) ÷ (NIR − red)
   • NIR × red
   • NIR − red without normalization
   • (NIR − red) ÷ (NIR + red)