Supervision: Adám Gosztolai

Project type: Semester project (master) Master thesis


Background Robotics and AI research has achieved significant feats in recent years so that current legged robots can perform a variety of tasks (c.f., Boston Dynamics robots). However, these feats result from substantial amount of training and are restricted to well-controlled environments. By contrast, insects can perform highly complex locomotor tasks right after birth, crucially, without the need for any learning (or training).

Project goals In this project, your job will be to distil principles of insect walking on a unique dataset that I spent the last year collecting in Prof Pavan Ramdya's lab. In this data, I recorded the fruit fly Drosophila melanogaster walking on terrains with variable difficulty while applying genetic perturbations to sensory neuron pathways. Because of these diverse perturbations, the dataset has the potential to unveil how insects adjust their neural circuits to cope with difficult terrains.

What can you learn? Your project will start by using the state-of-the-art LiftPose3D method to obtain 3D poses from the recordings from different camera angles. Then, you will gain experience building predictive models of robotic movement to discover the relevant latent variables that describe the changes in insect walking across environments and genetic perturbations. As an outcome, you will discover fundamental principles between the neural architecture the animals use and the locomotor dynamics they produce.

Outcome I expect this project to make a substantial impact in both neuroscience and robotic control and to be published in a high impact journal in 2023.

My background I am a mathematician (postdoc) with substantial experience in both neuroscience and machine learning (see more on my website). I have worked before in the labs of Pavan Ramdya and Auke Ijspeert and now I work with Pierre Vandergheynst.

Your background I am looking for someone with a quantitative profile (CS, EE, maths, robotics, data science etc). Biology knowledge is not necessary, but interest is appreciated.

If this project does not suit you, but you still want to work with me, get in touch and we will figure something out