Image Basics for Computer Vision Engineers
Image basics treat a picture as numeric data and build the first classifier on it. An image is an array of pixel values with height, width and colour channels. A linear classifier multiplies the flattened pixels by a weight matrix to produce class scores, trained with a loss such as softmax cross-entropy and gradient descent.
Stage: Foundations · Part of the Computer Vision Engineer roadmap · Last reviewed 19 November 2026
What to learn in Image Basics
- Load an image into a NumPy array and inspect its shape and value range
- Implement a k-nearest-neighbour classifier and tune k on a validation set
- Write a softmax loss and its gradient for a linear classifier
- Train with stochastic gradient descent and choose a learning rate
- Add L2 regularisation and describe its effect on the weights
Image Basics interview questions
Why do you need separate training, validation and test sets?
Training data fits the parameters, validation data is used to choose hyperparameters such as learning rate, and the test set gives a final unbiased estimate. If you tune on the test set, performance estimates become optimistic because the model indirectly adapts to it.
What does softmax do in a classifier?
Softmax turns raw class scores into positive numbers that sum to one, which can be read as class probabilities. Combined with cross-entropy loss, it penalises low probability on the correct class and gives a smooth gradient for training.
Where Image Basics fits in the Computer Vision Engineer roadmap
Image Basics belongs to the Foundations stage (intermediate level), which takes about 38 hours. Images and classifiers.
Other units in this stage
Back to the full Computer Vision Engineer roadmap