Your Instructor: Ass. Prof. Axel Carlier, University of Toulouse — Ass. Prof. Axel CARLIER, University of Toulouse
Monday Nov 28, 2016, 5:15pm-9:30pm Hanoi Time
General introduction to the course and the topic.
A few (already-known) administrative points can be found in the subsequent slides.
First part: General Introduction
Your first contrib (see also this example of short student Intro)
A few pictures to try existing Content-Based Image Retrieval websites:
requete10.jpg requete9.jpg requete8.jpg requete6.jpg requete3.jpg
Slides of the first class:
Tuesday Nov 29, 2016, 5:15pm-9:30pm Hanoi Time
Reminder on retrieval: boolean and vector models, evaluation
At the end of this lecture, you should have started thinking about your own dataset.
Wednesday Nov 30, 2016, 5:15pm-9:30pm Hanoi Time
Image retrieval with local features
At the end of this lecture, you should have started thinking about your own dataset.
Thursday Dec 01, 2016, 5:15pm-9:30pm Hanoi Time
Today is all about labs.
To start the labs, you will need to download the following archive: Code base for the labs
In order to make the code base work you need to:
Once you have installed these softwares, unpack the archive and run gradle wrapper in the extracted folder. Open the build.gradle file and change the paths to your test set and one of your query images. You can test the two sample codes that are provided:
Once the code runs, you can start implementing Precision/Recall measures to evaluate the quality of the retrieval. Use your Ground Truth to obtain the Precision/Recall on all query images and compute a global score of Precision and Recall for this algorithm, on all queries.
Friday Dec 02, 2016, 5:15pm-9:30pm Hanoi Time
Global descriptors, Bag of Words
In the mean time, we will continue the labs.