Class – 10th

50 multiple choice questions

Topic – computer vision
1. **What is the primary goal of computer vision?**

- A) Develop hardware for computers

- B) Enable computers to understand visual data

- C) Improve computer processing speed

- D) Create better user interfaces

2. **Which technique is used for detecting edges in an image?**

- A) Convolution

- B) Edge Detection

- C) Gaussian Blur

- D) Histogram Equalization

3. **What is the function of a convolutional layer in a Convolutional Neural Network

(CNN)?**

- A) Pooling features

- B) Activating neurons

- C) Detecting features

- D) Normalizing data

4. **Which algorithm is widely used for object detection?**

- A) R-CNN
 - B) SVM

- C) k-NN

- D) PCA

5. **What is the purpose of image segmentation?**

- A) Compress images

- B) Divide an image into meaningful parts

- C) Enhance image quality

- D) Convert images to grayscale

6. **Which of the following is a popular dataset used for image recognition?**

- A) MNIST

- B) CIFAR-10

- C) ImageNet

- D) All of the above

7. **In computer vision, what does the term 'feature extraction' refer to?**

- A) Reducing the image size

- B) Identifying important parts of the image

- C) Enhancing image contrast

- D) Converting images to vectors

8. **Which of the following is not a common color space in computer vision?**

- A) RGB

- B) CMYK
 - C) HSV

- D) HTTP

9. **What is the main advantage of using a CNN over a fully connected neural network
for image processing?**

- A) Faster processing

- B) Less prone to overfitting

- C) Better at handling spatial data

- D) Easier to implement

10. **What does the term 'pooling' refer to in the context of CNNs?**

- A) Combining multiple layers

- B) Down-sampling the input

- C) Normalizing the output

- D) Adding bias to neurons

11. **What is YOLO in computer vision?**

- A) A type of loss function

- B) An image compression technique

- C) A real-time object detection system

- D) A neural network architecture for image classification

12. **Which type of neural network is most commonly used for image processing tasks?
**

- A) Recurrent Neural Network (RNN)

 - B) Convolutional Neural Network (CNN)

- C) Feedforward Neural Network (FNN)

- D) Long Short-Term Memory (LSTM)

13. **Which method is used to reduce the dimensionality of image data?**

- A) Gradient Descent

- B) Principal Component Analysis (PCA)

- C) K-Means Clustering

- D) Support Vector Machines (SVM)

14. **In image classification, what does the 'Softmax' function do?**

- A) Normalizes image pixels

- B) Converts scores to probabilities

- C) Enhances image features

- D) Reduces image noise

15. **What is 'image augmentation' in the context of training neural networks?**

- A) Generating more data by transforming existing images

- B) Reducing the size of images

- C) Enhancing image quality

- D) Normalizing images

16. **What does 'CNN' stand for?**

- A) Convolutional Neural Network

- B) Categorical Neural Network

 - C) Convoluted Network Nodes

- D) Convergent Neural Network

17. **Which layer in a CNN is responsible for reducing the spatial size of the feature
maps?**

- A) Convolutional Layer

- B) Fully Connected Layer

- C) Pooling Layer

- D) Softmax Layer

18. **What is the purpose of using a rectified linear unit (ReLU) activation function?**

- A) To ensure non-linearity

- B) To reduce dimensions

- C) To classify images

- D) To detect edges

19. **What is 'backpropagation' used for in neural networks?**

- A) Initializing weights

- B) Forwarding inputs through the network

- C) Adjusting weights based on error

- D) Activating neurons

20. **Which of the following methods is used to handle overfitting in deep learning
models?**

- A) Increasing the learning rate

 - B) Data augmentation

- C) Reducing the training data

- D) Using fewer layers

21. **What is a 'filter' in the context of CNNs?**

- A) A matrix used for convolutions

- B) A technique for pooling

- C) A type of activation function

- D) A fully connected layer

22. **In computer vision, what does 'non-max suppression' help with?**

- A) Enhancing image quality

- B) Reducing false positives in object detection

- C) Normalizing pixel values

- D) Increasing the number of detected objects

23. **What is the role of an activation function in a neural network?**

- A) To activate neurons based on inputs

- B) To initialize weights

- C) To combine layers

- D) To normalize outputs

24. **Which of the following is a loss function commonly used for classification tasks?**

- A) Mean Squared Error (MSE)

- B) Cross-Entropy Loss
 - C) Hinge Loss

- D) Binary Cross-Entropy

25. **What does the term 'mean average precision (mAP)' refer to in object detection?**

- A) A metric for evaluating classification accuracy

- B) A technique for data augmentation

- C) A measure of the model’s detection accuracy

- D) A method for normalizing data

26. **Which algorithm is used for detecting corners in an image?**

- A) Harris Corner Detector

- B) SIFT

- C) HOG

- D) YOLO

27. **What is the purpose of dropout in neural networks?**

- A) To prevent overfitting

- B) To enhance activation functions

- C) To increase the learning rate

- D) To combine features

28. **Which of the following is a pre-trained model for image classification?**

- A) VGGNet

- B) AlexNet

- C) ResNet
 - D) All of the above

29. **What does 'ROI' stand for in object detection?**

- A) Region of Interest

- B) Rate of Increase

- C) Range of Intensity

- D) Radius of Influence

30. **Which algorithm is used for face detection in images?**

- A) Haar Cascades

- B) K-Means

- C) Random Forest

- D) Gradient Descent

31. **What does the term 'stride' refer to in CNNs?**

- A) The number of filters used

- B) The step size of the convolution filter

- C) The activation function applied

- D) The pooling operation performed

32. **What is the effect of increasing the kernel size in a convolutional layer?**

- A) Reduced spatial dimensions

- B) Increased computation cost

- C) More detailed feature detection

- D) Faster training time

33. **Which of the following methods is used for image classification?**

- A) R-CNN

- B) YOLO

- C) SIFT

- D) AlexNet

34. **In semantic segmentation, what is the main goal?**

- A) To classify images into categories

- B) To detect objects in images

- C) To assign a label to every pixel in an image

- D) To compress images

35. **What is the function of a 'deconvolution' layer in neural networks?**

- A) To reduce dimensions

- B) To perform convolution in reverse

- C) To pool features

- D) To activate neurons

36. **Which of the following techniques is used to enhance image contrast?**

- A) Histogram Equalization

- B) Smoothing

- C) Edge Detection

- D) Convolution
37. **What is the purpose of using a 'batch normalization' layer?**

- A) To normalize the inputs of each layer

- B) To increase training speed

- C) To reduce internal covariate shift

- D) All of the above

38. **Which model architecture is known for its skip connections?**

- A) VGGNet

- B) ResNet

- C) Inception

- D) AlexNet

39. **What is the purpose of an 'embedding' in deep learning?**

- A) To convert high-dimensional data to low-dimensional vectors

- B) To perform image segmentation

- C) To enhance image resolution

- D) To normalize pixel values

40. **Which of the following is a technique for reducing the number of parameters in a
CNN?**

- A) Data Augmentation

- B) Dropout

- C) Weight Sharing

- D) Batch Normalization
41. **What does the term 'transfer learning' refer to in the context of deep learning?**

- A) Training a model from scratch

- B) Using pre-trained models on new tasks

- C) Normalizing data before training

- D) Transferring data to a different format

42. **What is 'IoU' in the context of object detection?**

- A) Intersection over Union

- B) Input-output Unification

- C) Interpolation over Updates

- D) Initial Offset Unit

43. **Which of the following is a key challenge in computer vision?**

- A) Limited processing power

- B) Large amounts of data

- C) Variability in lighting and occlusions

- D) All of the above

44. **What does the 'dropout' technique do during training?**

- A) Increases the learning rate

- B) Drops some neurons randomly to prevent overfitting

- C) Normalizes the inputs

- D) Increases the number of layers

45. **Which neural network architecture is specifically designed to handle sequential
data?**

- A) CNN

- B) RNN

- C) GAN

- D) DBN

46. **What is the primary purpose of the activation function in a neural network?**

- A) To perform pooling operations

- B) To introduce non-linearity

- C) To initialize weights

- D) To perform convolutions

47. **Which of the following is a common pre-processing step in computer vision?**

- A) Normalizing pixel values

- B) Removing all colors

- C) Adding noise to the image

- D) Converting image to vector

48. **What does 'GAN' stand for in the context of neural networks?**

- A) Generalized Activation Network

- B) Generative Adversarial Network

- C) Gradient Ascend Network

- D) Graphical Adjacency Network

49. **What is the purpose of the fully connected layer in a CNN?**

- A) To perform convolutions

- B) To connect every neuron in one layer to every neuron in another layer

- C) To pool features

- D) To normalize inputs

50. **Which of the following optimizers is commonly used in training deep learning
models?**

- A) Stochastic Gradient Descent (SGD)

- B) Adam

- C) RMSprop

- D) All of the above

Answer

1. **B) Enable computers to understand visual data**

2. **B) Edge Detection**

3. **C) Detecting features**

4. **A) R-CNN**

5. **B) Divide an image into meaningful parts**

6. **D) All of the above**

7. **B) Identifying important parts of the image**

8. **D) HTTP**

9. **C) Better at handling spatial data**

10. **B) Down-sampling the input**

11. **C) A real-time object detection system**

12. **B) Convolutional Neural Network (CNN)**

13. **B) Principal Component Analysis (PCA)**

14. **B) Converts scores to probabilities**

15. **A) Generating more data by transforming existing images**

16. **A) Convolutional Neural Network**

17. **C) Pooling Layer**

18. **A) To ensure non-linearity**

19. **C) Adjusting weights based on error**

20. **B) Data augmentation**

21. **A) A matrix used for convolutions**

22. **B) Reducing false positives in object detection**

23. **A) To activate neurons based on inputs**

24. **B) Cross-Entropy Loss**

25. **C) A measure of the model’s detection accuracy**

26. **A) Harris Corner Detector**

27. **A) To prevent overfitting**

28. **D) All of the above**

29. **A) Region of Interest**

30. **A) Haar Cascades**

31. **B) The step size of the convolution filter**

32. **C) More detailed feature detection**

33. **D) AlexNet**

34. **C) To assign a label to every pixel in an image**

35. **B) To perform convolution in reverse**

36. **A) Histogram Equalization**

37. **D) All of the above**

38. **B) ResNet**

39. **A) To convert high-dimensional data to low-dimensional vectors**

40. **C) Weight Sharing**

41. **B) Using pre-trained models on new tasks**

42. **A) Intersection over Union**

43. **D) All of the above**

44. **B) Drops some neurons randomly to prevent overfitting**

45. **B) RNN**

46. **B) To introduce non-linearity**

47. **A) Normalizing pixel values**

48. **B) Generative Adversarial Network**

49. **B) To connect every neuron in one layer to every neuron in another layer**

50. **D) All of the above**