Developing Kubeflow Pipelines: Kaggle’s Facial Keypoints Detection Competition

Detect the location of keypoints on face images with machine learning.

Date: Aug 24, 2022 06:00 AM Pacific Time (US and Canada)
Kubeflow familiarity: Beginner to Intermediate

About the Workshop:
In this workshop we’ll show how to turn Kaggle’s Facial Keypoints Detection competition into a Kubeflow Pipeline using the KFP SDK and the Kale JupyterLab extension.

About the Kaggle Competition:
The objective of the Facial Keypoints Detection Kaggle competition, as the site notes, is to predict keypoint positions on face images. This can be used as a building block in several applications, such as:

  • Tracking faces in images and video
  • Analyzing facial expressions
  • Detecting dysmorphic facial signs for medical diagnosis
  • Biometrics/face recognition
In the world of machine learning, it is well known that detecting facial keypoints is a very difficult problem to solve. This is because facial features vary significantly from person to person, and even for a specific individual, there is a large amount of variation in facial images due to changing conditions such as 3D pose, size, position, viewing angle, and lighting. Although computer vision research has come a long way in addressing these difficulties, there still remain many opportunities for improvement!

Register Now!

Upcoming virtual workshops