DNN-based face detection in Computer Vision - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to load a pre-trained DNN face detector model.

Computer Vision

net = cv2.dnn.readNetFromCaffe('[1]', 'deploy.caffemodel')

Drag options to blanks, or click blank then click option'

Aconfig.json

Bmodel.pb

Cdeploy.prototxt

Dweights.h5

Attempts:

3 left

💡 Hint

Common Mistakes

Using the weights file instead of the prototxt file for architecture.

Confusing TensorFlow or Keras model files with Caffe files.

The prototxt file defines the model architecture for the Caffe DNN face detector.

2fill in blank

medium

Complete the code to prepare the input blob for the DNN face detector.

Computer Vision

blob = cv2.dnn.blobFromImage(image, 1.0, ([1], [1]), (104.0, 177.0, 123.0))

Drag options to blanks, or click blank then click option'

A300

B224

C128

D512

Attempts:

3 left

💡 Hint

Common Mistakes

Using incorrect input size causing detection failure.

Confusing input size with image resolution.

The face detector expects 300x300 input size for the blob.

3fill in blank

hard

Fix the error in the code to extract confidence scores from the detections.

Computer Vision

confidence = detections[0, 0, i, [1]]

Drag options to blanks, or click blank then click option'

A1

B0

C3

D2

Attempts:

3 left

💡 Hint

Common Mistakes

Using index 1 or 3 which correspond to class or bounding box coordinates.

Confusing zero-based indexing.

Index 2 holds the confidence score in the detections array.

4fill in blank

hard

Fill both blanks to compute bounding box coordinates from detections.

Computer Vision

box = detections[0, 0, i, 3:7] * np.array([[1], [2], [1], [2]])

Drag options to blanks, or click blank then click option'

Aw

Bh

Cimage.shape[1]

Dimage.shape[0]

Attempts:

3 left

💡 Hint

Common Mistakes

Swapping width and height values.

Using fixed numbers instead of image dimensions.

Width is image.shape[1] and height is image.shape[0], used to scale box coordinates.

5fill in blank

hard

Fill all three blanks to filter detections by confidence and draw rectangles.

Computer Vision

if confidence > [1]:
    (startX, startY, endX, endY) = box.astype([2])
    cv2.rectangle(image, (startX, startY), (endX, endY), [3], 2)

Drag options to blanks, or click blank then click option'

A0.5

Bint

C(0, 255, 0)

Dfloat

Attempts:

3 left

💡 Hint

Common Mistakes

Using too low or too high confidence threshold.

Not converting coordinates to int causing errors.

Using wrong color format for rectangle.

Confidence threshold is 0.5, box coordinates converted to int, green rectangle color.