Introduction

This short guide describes core concepts of the product, shows main FaceEngine features and suggests usage scenarios.

This document is not a full-featured API reference manual nor a step by step tutorial. For reference pages, please see Doxygen API documentation that is shipped with FaceEngine. For complete examples, please head to our developer portal.

What this book does, however, is this:

• It describes ideas behind resource management and gives a clue why one or another decision was made. With this in mind, you are ready to write efficient code with FaceEngine;

• It breaks down full face analysis and recognition pipeline in parts and shows how one part affects all the others. This information will help you to adapt FaceEngine to your needs, which is somewhat more productive than blindly following tutorials;

• It details things that are important and omits things that are obvious, so you get information that matters most.

This book is split up into several chapters. There are chapters dedicated to each FaceEngine facility; there are chapters with conceptual overviews; there are chapters with generic information. We tried to write the book starting from low-level concepts and moving on to face detection, description and recognition tasks solving one problem at a time. Although sometimes we just had to give references to chapters ahead, we tried to minimize such jumps.

The opening chapter of this book is called "Core concepts". It will tell you about memory management techniques, object creation and destruction strategies that are widely used across the entire FaceEngine. The following chapters catch up telling how higher level FaceEngine components are created from those building blocks.