Why detection localizes objects in PyTorch - Test Your Understanding

Practice - 5 Tasks

1fill in blank

easy

Complete the code to create a bounding box tensor for an object detection model.

PyTorch

bbox = torch.tensor([[1], 50, 150, 200])

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

A30

B'30'

C(30)

D[30]

Attempts:

3 left

💡 Hint

Common Mistakes

Using quotes around numbers, which makes them strings.

Using list or tuple instead of a number.

The bounding box coordinates must be numeric values, so 30 without quotes is correct.

2fill in blank

medium

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

PyTorch

iou = intersection_area / [1]

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

Abox1_area

Bintersection_area

Cunion_area

Dbox2_area

Attempts:

3 left

💡 Hint

Common Mistakes

Dividing by intersection area instead of union area.

Using only one box's area.

IoU is the ratio of intersection area over union area of two boxes.

3fill in blank

hard

Fix the error in the code that extracts predicted bounding boxes from model output.

PyTorch

pred_boxes = outputs['boxes'][1]

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

A(0)

B[0]

C{0}

D<0>

Attempts:

3 left

💡 Hint

Common Mistakes

Using parentheses instead of brackets for indexing.

Using curly braces or angle brackets.

To index a tensor in PyTorch, square brackets [] are used, so [0] is correct.

4fill in blank

hard

Fill both blanks to complete the code that filters detections by confidence score.

PyTorch

filtered = [box for box in detections if box['score'] [1] [2]]

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

A>

B0.5

C<

D1.0

Attempts:

3 left

💡 Hint

Common Mistakes

Using less than instead of greater than.

Using threshold 1.0 which is too strict.

We keep boxes with score greater than 0.5 to filter confident detections.

5fill in blank

hard

Fill all three blanks to complete the code that creates a dictionary of detected object labels and their bounding boxes.

PyTorch

results = [1]((label, [2]) for label, box in zip(labels, [3]))

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

Adict

Bbox

Cboxes

Dlist

Attempts:

3 left

💡 Hint

Common Mistakes

Using list instead of dict.

Using wrong variable names for boxes.

We use dict() to create a dictionary mapping each label to its bounding box from the boxes list.