“Life keeps throwing me stones. And I keep finding the diamonds.”– Ana Claudia Antunes

Hi enthusiastic people, today I am working in a mine of Deep Learning. Read my post to know more.

If you are still following me, it means that you are a fan of Content-Based Image Retrieval (CBIR). So, let’s talk about the mentioned challenges (in the previous post). The first and the most important part of having an optimum search engine is finding a proper Feature Extractor (FE). So, after writing the whole pip- line, we need to define our FE. But a question would appear here, WHICH FE WOULD BE THE BEST? answering this question depends on the aim of using FE and the data set that we are going to consider. As we are all aware, our data set is Whole Slide Images (WSIs) and they are gigapixel. To work with these big-size images we need to split them into different patches, then, extract the most meaningful features from them. This big size is now the stones on my way. During all past days, I was working on different approaches to tackle the first challenge to extract the significant features of our WSIs. To do so, I registered for a beneficial summer school was held on by Graz University of Technology (TU Graz) and it gave me some brainstorms.

In addition, I moved to Norway, Stavanger to participate with our other ESRs who work at Universitetet I Stavanger (UIS). This is my second secondment.


Figure 1: A picture of UIS

 Figure 2: The Flørli stair with 4.444 steps is one of the longest wooden staircases in the world.  Sep 26, 2021.  Flørli, Norway. Zahra (ESR9) with Umay (ESR11) and other Norwegian friends.


So, now I am mining to find my diamond. If you want to know the results of these brainstorms and collaborations just follow our post.


Zahra Tabatabei – ESR9.