Description
Orthogonal Research and Education Laboratory is seeking to further develop the Representational Brains and Phenotypes project by building scientific simulations for deployment in basic research and educational outreach. We aim to construct a dynamical simulation involving several types of Braitenberg vehicle as described by the neuroscientist Valentino Braitenberg in his classic book "Vehicles: experiments in synthetic Psychology" (1984). Come be a part of bringing these thought experiments to life!
Aims
These simulations will involve building a virtual model of a vehicle, complete with the body, sensors, effectors, and internal neural wiring. Simulating changes in the wiring based on environmental stimuli and bodily constraints is the most important part of the model, as we seek to go beyond Braitenberg’s original vision and incorporate the construction of highly complex nervous systems (neuronal networks). As an applicant, you will propose a way to create a robust and efficient simulation. The project mentor will provide extensive feedback throughout the process.
The first part of the project will involve constructing a simulation of the vehicles, while the second part will involve a simulation of the neuronal network (formation, consolidation, and deletion of connections). A successful project will propose a method for neuronal network simulation, whether that be a neural network, a genetic algorithm, or a hybrid model. The finished product will be used for the study of animal connectomes, educational simulations, and guiding physical robot design.
Skills
Open source development experience with C++, Python, Java. An interest in the underlying biological processes is essential. Applicants with strong abstract thinking abilities are preferred, as is previous background and/or interest in Cognitive Science research. Good communication skills and familiarity with open science practices are expected.
Hello, Neurostars! I am a previous GSoC student ('17 working with EBI), currently studying Bioinformatics and have a strong interest in Cognitive Science.
Now, the project sounds really interesting and I’m already starting to think about how this could be implemented. Herefore, it would be interesting to look at already existing cognitive simulations - maybe you know one you find particulary well-coded or efficient? Would be glad to hear any suggestions though!
Thanks in advance, Stefan.
Hello There!
I am a third-year engineering student from Indian Institute of Technology (IIT), Kharagpur. I have been going through the GSoC’19 project descriptions at INCF, and the project named “Modeling Neural Development with Braitenberg Vehicles” looks most interesting to work on for me. As a follow-up, I’m halfway to the book “Vehicles: Experiments in Synthetic Psychology” by Valentino Braitenberg, to understanding the underlying concepts and I am really enjoying it.
I have been inclined to this field thanks to the course named “Introduction to Cognitive Information Processing”, which I audited last semester at my University. So the willingness to work on something like this was always there. Also, being a programmer I wanted to work on something exactly like this which aligns my skills and interest both.
So, I humbly request you to guide me further about this project. I would be very happy to get assistance from you about things I can do/read or work on to get a more clear picture of this project. Given my credentials, I hope that I am qualified enough to work on this and would be able to work on it with your guidance. Looking forward to hearing back sooner!
If you are interested in an existing set of models/modeling framework that might work well with this project, check out SimBrain: http://www.simbrain.net/
You can implement Neural Network-style models directly in the interface, and there is even a traditional implementation of Briatenberg vehicles: https://www.youtube.com/watch?v=OalJO1C_cHQ
This does not include the developmental component, but building off this demo might be a shorter path to success. Not sure how it integrates with modules for genetic algorithms, etc.