How To Apply Computer Vision to Build an Emotion-Based Dog Filter in Python 3

The author selected Girls Who Code to receive a donation as part of the Write for DOnations program.

Introduction

Computer vision is a subfield of computer science that aims to extract a higher-order understanding from images and videos. This field includes tasks such as object detection, image restoration (matrix completion), and optical flow. Computer vision powers technologies such as self-driving car prototypes, employee-less grocery stores, fun Snapchat filters, and your mobile device’s face authenticator.

In this tutorial, you will explore computer vision as you use pre-trained models to build a Snapchat-esque dog filter. For those unfamiliar with Snapchat, this filter will detect your face and then superimpose a dog mask on it. You will then train a face-emotion classifier so that the filter can pick dog masks based on emotion, such as a corgi for happy or a pug for sad. Along the way, you will also explore related concepts in both ordinary least squares and computer vision, which will expose you to the fundamentals of machine learning.

👁 A working dog filter

As you work through the tutorial, you’ll use OpenCV, a computer-vision library, numpy for linear algebra utilities, and matplotlib for plotting. You’ll also apply the following concepts as you build a computer-vision application:

• Ordinary least squares as a regression and classification technique.
• The basics of stochastic gradient neural networks.

While not necessary to complete this tutorial, you’ll find it easier to understand some of the more detailed explanations if you’re familiar with these mathematical concepts:

• Fundamental linear algebra concepts: scalars, vectors, and matrices.
• Fundamental calculus: how to take a derivative.

You can find the complete code for this tutorial at https://github.com/do-community/emotion-based-dog-filter.

Let’s get started.