Masked and unmasked images

Course Q&A Deep Learning Specialization Convolutional Neural Networks
1

Hello everyone

I have a very basic question. What is the difference between masked and unmasked images? Can someone explain in terms of car detection. What is the significance of using masked and unmasked images in U-net? How it is helpful in car detection?

Thanks
Arunima

2

Hey @arunimash28 I couldn’t find car detection related image. But this image illustrates the point just as same.

On the left you have the actual image. The unmasked image. Meaning, speaking in terms of classification, this is a raw picture where no chairs have been identified.

On the right is the masked image. Meaning, speaking in terms of classification, the the objects which has been classified as chairs have been covered with this blue colour mask.
Screenshot 2021-07-27 at 2.58.00 AM

4

Thank you so much for your response. In the assignment for U-NET , both the masked and unmasked image are being uploaded and then the predicted image has been compared to the masked image. My next question is how to generate the masked image from the unmasked image ?

5

Hey @arunimash28, hmmm, I am not exactly sure about that. But let me search on this and I’ll get back to you in some time.