Keras Preprocessing

[Thomi Aditya Alhakiim]

I got this paper titled 'Deep Contrast Learning for Salient Object Detection' by Guanbin LI, Yizhou Yu a few day ago and I tried to implement the method using keras modul API (Im new to deep learning btw). I tried the first stream one that using CNN and I think I nailed it, but the second stream needed spatial pooling using superpixel segmentation image to implement. How can I convert the image to superpixel on preprocessing before I get this in to the model? Or is any other method I can use to answer my problem?

The first image above shows the two streams used for the method to work. The second image shows the Architecture for the CNN on the first stream.

Thanks.

[Sayak Paul]

The following two links might be helpful:

• https://keras.io/examples/vision/super_resolution_sub_pixel
• https://www.tensorflow.org/addons/api_docs/python/tfa/layers/SpatialPyramidPooling2D

Sayak Paul | sayak.dev

--
You received this message because you are subscribed to the Google Groups "Keras-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to keras-users...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/keras-users/102402be-bcf1-42a5-aff2-3b388f05d6c4n%40googlegroups.com.