Image Segmentation with Mask R-CNN

Picture Source: Doğu İlmak

Content

You will see how to use a Mask R-CNN model from TensorFlow Hub for object detection and instance segmentation. This means that aside from the bounding boxes, the model is also able to predict segmentation masks for each instance of a class in the image. In this repository, we are going to use photos that I selected from the photos I took in the cities I visited. You can use my pictures of the cities (Berlin, Budapest, Gdańsk, Kraków, Olsztyn, Oslo, Prague and Warsaw) for image segmentation. In addition, I would like to thank Laurence Moroney. I gained most of the knowledge I have gained thanks to him. I have endless respect to you.

References

1. Laurence Moroney
2. Advanced Computer Vision with TensorFlow
3. TensorFlow GitHub

Keywords

• Mask R-CNN
• Image Segmentation
• Instance Segmentation
• Computer Science
• Inception ResNet V2 1024x1024

Segmentation

Image Without Segmentation

Image With Segmentation

Contact Me

If you have something to say to me please contact me:

• Twitter: Doguilmak
• Mail address: doguilmak@gmail.com