You're reading from Hands-On Computer Vision with Detectron2

Product type Book

Published in Apr 2023

Publisher Packt

ISBN-13 9781800561625

Pages 318 pages

Edition 1st Edition

Languages

Python

Concepts

Computer Vision

Author (1):

Van Vung Pham

Table of Contents (20) Chapters

Preface

1. Part 1: Introduction to Detectron2

2. Chapter 1: An Introduction to Detectron2 and Computer Vision Tasks

3. Chapter 2: Developing Computer Vision Applications Using Existing Detectron2 Models

4. Part 2: Developing Custom Object Detection Models

5. Chapter 3: Data Preparation for Object Detection Applications

6. Chapter 4: The Architecture of the Object Detection Model in Detectron2

7. Chapter 5: Training Custom Object Detection Models

8. Chapter 6: Inspecting Training Results and Fine-Tuning Detectron2’s Solvers

9. Chapter 7: Fine-Tuning Object Detection Models

10. Chapter 8: Image Data Augmentation Techniques

11. Chapter 9: Applying Train-Time and Test-Time Image Augmentations

12. Part 3: Developing a Custom Detectron2 Model for Instance Segmentation Tasks

13. Chapter 10: Training Instance Segmentation Models

14. Chapter 11: Fine-Tuning Instance Segmentation Models

15. Part 4: Deploying Detectron2 Models into Production

16. Chapter 12: Deploying Detectron2 Models into Server Environments

17. Chapter 13: Deploying Detectron2 Models into Browsers and Mobile Environments

18. Index

Why subscribe?

19. Other Books You May Enjoy

Inspecting Training Results and Fine-Tuning Detectron2’s Solvers

This chapter covers how to use TensorBoard to inspect training histories. It utilizes the code and visualization approach to explain the concepts behind Detectron2’s solvers and their hyperparameters. The related concepts include gradient descent, stochastic gradient descent, momentum, and variable learning rate optimizers. This chapter also provides code to help you find the standard hyperparameters for Detectron2’s solvers.

By the end of this chapter, you will be able to use TensorBoard to analyze training results and find insights. You will also have a deep understanding of the essential hyperparameters for Detectron2’s solvers. Additionally, you will be able to use code to generate appropriate values for these hyperparameters on your custom dataset. Specifically, this chapter covers the following topics: