From the course: Hands-On AI: Computer Vision Projects with Ultralytics and OpenCV
Unlock this course with a free trial
Join today to access over 24,900 courses taught by industry experts.
Validate the model - OpenCV Tutorial
From the course: Hands-On AI: Computer Vision Projects with Ultralytics and OpenCV
Validate the model
- [Instructor] Let's say you have spent a lot of time on some computer VM project where you have created the data set, created the model, and now you want to deploy the model into the production. Without validation, you cannot deploy the model. Validation is a critical step in the machine learning pipeline that allow us to access the quality of our trained model. In this video, we will explore how we can use the validation mode with the Ultralytics Python package to validate different models that we have created. All the resources used in this video are available in the course code subdirectory 03-06. In the subdirectory, you can see two different models. That is the model that we have created in video two of this chapter with the help of object detection to detect apples. best-segment.pt is the model that we have created in video four of this chapter with the help of image segmentation model for segmentation of apples. We will use these models here and validate them. Let's take a…
Practice while you learn with exercise files
Download the files the instructor uses to teach the course. Follow along and learn by watching, listening and practicing.
Contents
-
-
-
-
-
Overview of Ultralytics tasks and modes14m 52s
-
(Locked)
Training an object-detection model and inference16m 54s
-
(Locked)
Auto-annotate detection data to segmentation format9m 10s
-
(Locked)
Training and inference for an image-segmentation model11m 31s
-
(Locked)
How to use the pose estimation model7m 48s
-
(Locked)
Validate the model9m 16s
-
(Locked)
How to use other computer vision models11m 20s
-
(Locked)
How to predict and track the detected objects11m 13s
-
(Locked)
How to benchmarks different models9m 28s
-
(Locked)
Export models to different formats8m 35s
-
-
-