Annotation quality in Computer Vision - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to calculate the Intersection over Union (IoU) between two bounding boxes.

Computer Vision

def iou(boxA, boxB):
    xA = max(boxA[0], boxB[0])
    yA = max(boxA[1], boxB[1])
    xB = min(boxA[2], boxB[2])
    yB = min(boxA[3], boxB[3])

    interArea = max(0, xB - xA) * max(0, yB - yA)

    boxAArea = (boxA[2] - boxA[0]) * (boxA[3] - boxA[1])
    boxBArea = (boxB[2] - boxB[0]) * (boxB[3] - boxB[1])

    iou = interArea / [1]
    return iou

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

AboxAArea + boxBArea - interArea

BboxAArea - boxBArea

CinterArea

DboxAArea + boxBArea

Attempts:

3 left

💡 Hint

Common Mistakes

Using only the sum of areas without subtracting the intersection area.

Dividing by intersection area instead of union area.

The IoU is the ratio of the intersection area over the union area. The union area is calculated as the sum of both box areas minus the intersection area.

2fill in blank

medium

Complete the code to compute the precision of an object detection model given true positives and false positives.

Computer Vision

def precision(true_positives, false_positives):
    return true_positives / [1]

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

Afalse_positives

Btrue_positives + false_negatives

Ctrue_positives + false_positives

Dtrue_positives

Attempts:

3 left

💡 Hint

Common Mistakes

Using false negatives in the denominator instead of false positives.

Dividing by only true positives.

Precision is the ratio of true positives over all predicted positives (true positives + false positives).

3fill in blank

hard

Fix the error in the code to calculate recall given true positives and false negatives.

Computer Vision

def recall(true_positives, false_negatives):
    return true_positives / [1]

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

Afalse_negatives

Btrue_positives + false_negatives

Ctrue_positives + false_positives

Dtrue_positives

Attempts:

3 left

💡 Hint

Common Mistakes

Using false positives in the denominator instead of false negatives.

Dividing by only true positives.

Recall is the ratio of true positives over all actual positives (true positives + false negatives).

4fill in blank

hard

Fill both blanks to create a dictionary comprehension that maps image IDs to their annotation counts, but only for images with more than 5 annotations.

Computer Vision

annotation_counts = {img_id: [1] for img_id, annotations in dataset.items() if [2]

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

Alen(annotations)

Blen(annotations) > 5

Clen(annotations) < 5

Dannotations

Attempts:

3 left

💡 Hint

Common Mistakes

Using the annotations list directly instead of its length.

Using the wrong comparison operator in the if condition.

The dictionary comprehension counts annotations per image and filters to include only those with more than 5 annotations.

5fill in blank

hard

Fill all three blanks to create a list comprehension that extracts labels from annotations only if the confidence score is above 0.8.

Computer Vision

high_conf_labels = [ann[1] for ann in annotations if ann[2] > [3]]

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

A['label']

B['confidence']

C0.8

D['score']

Attempts:

3 left

💡 Hint

Common Mistakes

Using the wrong key for confidence score.

Using dot notation instead of square brackets for dictionary keys.

Using a wrong threshold value.

The list comprehension extracts the 'label' from each annotation where the 'confidence' score is greater than 0.8.