User Interface for an image processing pipeline that generates images useful for experiments in AI/Neuroscience
Organization
Computational Neuroscience Group
Abstract
To generate images that hide the object in an image in form of a star constellation, we have developed an image processing pipeline at the CNS group. This dataset currently generated using this pipeline is being used to study the computational notion of concepts and the process of image understanding in humans. The input and the intermediate stage images look like Figure 1 (see pdf). Few manual steps are still required in this process to select a particular setting or clean the outline version of the image with black colour spread tool in image editing software. The objective of this thesis is to create a software tool that allows the user to make these choices in a single UI. The software tool will be published along with the release of the dataset.
Graduation Theses defence year
2021-2022
Supervisor
Tarun Khajuria, Jaan Aru
Spoken language (s)
English
Requirements for candidates
Ability to build a UI on top of existing python code. It will desirable to have a working understanding of Image handling and processing tools.
Level
Bachelor
Keywords
Application of contact
Name
Tarun Khajuria
Phone
E-mail
Full Document