Why computer vision teaches machines to see - Challenge Your Understanding

Challenge - 5 Problems

🧠 Conceptual

intermediate

What is the main goal of computer vision?

Computer vision helps machines understand images and videos. What is the main goal of teaching machines to see?

A To allow machines to write text without any visual input

B To make machines generate random images without understanding

C To teach machines to hear sounds from images

D To enable machines to recognize and interpret visual information like humans do

💡 Hint

Think about what seeing means for humans and how machines might do the same.

Computer vision aims to help machines recognize and interpret visual data, similar to how humans see and understand the world.

❓ Predict Output

intermediate

Output of image pixel normalization code

What is the output of this Python code that normalizes pixel values of an image array?

Computer Vision

import numpy as np
image = np.array([[0, 128], [255, 64]])
normalized = image / 255
print(normalized)

A TypeError: unsupported operand type(s) for /: 'list' and 'int'

B [[0, 0.5], [1, 0.25]]

C [[0.0, 0.50196078], [1.0, 0.25098039]]

D [[0, 128], [255, 64]]

💡 Hint

Division by 255 converts pixel values to a 0-1 range as floats.

Dividing the numpy array by 255 converts each pixel to a float between 0 and 1. The exact float values are decimals, not rounded.

❓ Model Choice

advanced

Best model type for object detection in images

You want to build a system that finds and labels objects in photos. Which model type is best suited for this task?

A Convolutional Neural Network (CNN) with region proposal networks

B Recurrent Neural Network (RNN) for sequence prediction

C Linear Regression model for numeric prediction

D K-Means clustering for grouping unlabeled data

💡 Hint

Object detection needs spatial understanding of images.

CNNs with region proposal networks like Faster R-CNN are designed to detect and classify objects within images effectively.

❓ Metrics

advanced

Choosing the right metric for image classification accuracy

You trained a model to classify images into categories. Which metric best shows how often the model predicts the correct category?

A Mean Squared Error

B Accuracy

C Silhouette Score

D Perplexity

💡 Hint

Think about a metric that measures correct predictions over total predictions.

Accuracy measures the proportion of correct predictions out of all predictions, ideal for classification tasks.

🔧 Debug

expert

Why does this image preprocessing code raise an error?

What error does this code raise and why? import cv2 image = cv2.imread('photo.jpg') gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY) resized = cv2.resize(gray, (100, 100)) print(resized.shape)

Computer Vision

import cv2
image = cv2.imread('photo.jpg')
gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
resized = cv2.resize(gray, (100, 100))
print(resized.shape)

A AttributeError because 'image' is None if file not found

B TypeError because cv2.resize expects a list, not a numpy array

C ValueError because color conversion code is invalid

D No error, prints (100, 100)

💡 Hint

Check if the image file was loaded correctly before processing.

If 'photo.jpg' is missing or path is wrong, cv2.imread returns None. Passing None to cvtColor causes an error.