Data Skeptic

k-means Image Segmentation

Feb 22, 2022

23:01

Creator website

Highlights

AI Chapters

Episode notes

Using Weights and Biase Platform for Machine Learning

04:22

Using Unsupervised Learning Algorithms to Make Differential Groups?

01:40

Speck Rom

01:39

Introduction

4min

Speck Rom

2min

Is There a Difference Between the Foreground and the Background?

2min

How Long Does It Take for a Bird to Come Into Focus?

2min

Using Weights and Biase Platform for Machine Learning

4min

Using Unsupervised Learning Algorithms to Make Differential Groups?

2min

Is There Any Worth While Pursuit in Changing the Loss Function?

2min

Camouflage - A Parrot in a Bush Hat

2min

What's the Best Quality Photo?

2min

10.

Can We Cluster People?

2min

Linh Da joins us to explore how image segmentation can be done using k-means clustering. Image segmentation involves dividing an image into a distinct set of segments. One such approach is to do this purely on color, in which case, k-means clustering is a good option.

Check out our website for extended show notes and images! Thanks to our Sponsors: Visit Weights and Biases mention Data Skeptic when you request a demo! & Nomad Data In the image below, you can see the k-means clustering segmentation results for the same image with the values of 2, 4, 6, and 8 for k.