<div dir="ltr"><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">Beating Roger Federer: </div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">Modeling visual learning and expertise through a bioinspired neural network embedded in an electronic device</div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial"><br></div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">Goal: PhD grant. </div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">When: starting from September 2018 (date may be flexible to some extent)</div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">Topic: How do expert tennis players, like Roger Federer for example, predict if a ball will bounce in or out the field to decide if it should be played or not? After thousands of trajectory presentations, best champions have developed extraordinary skills in such a task, but little is known on how the visual system turns selective to spatiotemporal properties of the visual stimulus (e.g., 3D position, velocity and acceleration) and learns how to make an efficient use of it.</div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">The goal of the project is to build an embedded system – based on FPGA circuits and ARM processor – an artificial neural network which would replicate – and perhaps beat – the visual and anticipatory performances of these expert players. </div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">To achieve this goal successfully, we will develop a bio-inspired neural network, based on some of the key properties of human vision: the Smart NeuroCam (GST company) will be used to reproduce the retina functioning. It triggers its message under the form of spikes, in an asynchronous way (without any concept of frame per second), responding to spatial or temporal changes in the pattern of illumination. Several kinds of pre-processing filters can be implemented in VHDL language directly in the FPGA circuits, and the output is then sent to a neural network. The artificial network will learn to use this message, applying a simple learning rule, the Spike-Timing-Dependent Plasticity (STDP). This rule allows each neuron to become selective to a particular property of the stimulus, completely autonomously and with no supervision. Several layers will be built to allow perceiving more and more complex properties of the visual scene. Once the network will be established, its performances will be assessed in different conditions of learning and compared to those of the best tennis players.</div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial"><br></div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">The project is funded by a French National Research Agency (ANR), involving two sites and several researchers: </div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">−<span style="white-space:pre-wrap"> </span>Robin Baurès, Benoit Cottereau, Timothée Masquelier and Simon Thorpe, CerCo, Toulouse.</div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">−<span style="white-space:pre-wrap"> </span>Michel Paindavoine, GST, Dijon.</div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial"><br></div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">Where: The candidate will be based at CerCo, Toulouse (France), and will make the interface with the two sites, with regular trips. The computational neuroscience part will be done at Toulouse, and electronic part at Dijon. </div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial"><br></div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">Tasks:</div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">−<span style="white-space:pre-wrap"> </span>Matlab (or Python) based simulations of numerical filters. These filters will be applied to the image processing from which spikes are generated and then sent to feed the neural network and STDP learning mechanism.</div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">−<span style="white-space:pre-wrap"> </span>VHDL coding to implement these numerical filters into the FPGA circuits of the cameras</div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">−<span style="white-space:pre-wrap"> </span>C/C++ coding of the neural network and STDP mechanism that should work on an embedded ARM processor system</div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">−<span style="white-space:pre-wrap"> </span>Experimental tests that will allow evaluating the performance of the whole system, from spikes generation to visual properties learning of the embedded system, to predict tennis ball’s trajectories </div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial"><br></div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">Required skills: </div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">−<span style="white-space:pre-wrap"> </span>Strong knowledge on electronic, and openness to computational neurosciences</div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">−<span style="white-space:pre-wrap"> </span>Knowledge in signal-image processing, and artificial neural network</div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">−<span style="white-space:pre-wrap"> </span>Interest for multidisciplinary research</div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">−<span style="white-space:pre-wrap"> </span>Ability to turn smoothly autonomous, once the road has been set</div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">−<span style="white-space:pre-wrap"> </span>Ability to be at the interface of two scientific fields and two working areas</div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">−<span style="white-space:pre-wrap"> </span>Programming with Matlab and/or Python for simulating the neural network</div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">−<span style="white-space:pre-wrap"> </span>Programming in VHDL language for FPGA circuits</div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">−<span style="white-space:pre-wrap"> </span>Programming in C/C++ language for porting the algorithms on ARM microprocessors</div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">−<span style="white-space:pre-wrap"> </span>French is not a requirement if fluent in English, but willingness to learn would be beneficial </div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial"><br></div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">Citizenship Eligibility: For security / defense reasons, an evaluation will be made of the candidates before any possible acceptation. </div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial"><br></div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">Relevant publications for the project:</div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">−<span style="white-space:pre-wrap"> </span>Masquelier, T., Guyonneau, R. & Thorpe S.J. (2009). Competitive STDP-Based Spike Pattern Learning. Neural Comput, 21(5),1259-1276.</div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">−<span style="white-space:pre-wrap"> </span>Masquelier, T. & Thorpe, S.J. (2007). Unsupervised learning of visual features through spike timing dependent plasticity. PLoS Comput Biol, 3(2):e31.</div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">−<span style="white-space:pre-wrap"> </span>Cottereau, B.R., McKee, S.P. & Norcia, A.M. (2014). Dynamics and cortical distribution of neural responses to 2D and 3D motion in human. Journal of Neurophysiology 111(3), 533-543.</div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">−<span style="white-space:pre-wrap"> </span>SmartNeuroCam de GST :<span> </span><a href="https://gsensing.eu/fr/category/sections/products" target="_blank" style="color:rgb(17,85,204)">https://gsensing.eu/fr/<wbr>category/sections/products</a></div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial"><br></div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">Contact:</div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">Robin Baurès, PhD</div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">Associate Professor</div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">CerCo, Université Toulouse 3, CNRS</div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">CHU Purpan, Pavillon Baudot</div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">31059 Toulouse Cedex 9 – France</div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">Office phone: 0033 (0)5 62 74 62 15</div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">Email :<span> </span><a href="mailto:robin.baures@cnrs.fr" target="_blank" style="color:rgb(17,85,204)">robin.baures@cnrs.fr</a></div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial"><br></div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">Pr Michel Paindavoine</div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">GlobalSensing Technologies</div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">14, rue Pierre de Coubertin</div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">21000 Dijon</div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">email :<span> </span><a href="mailto:michel.paindavoine@gsensing.eu" target="_blank" style="color:rgb(17,85,204)">michel.paindavoine@gsensing.eu</a></div><div style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial"><br></div>-- <br><div class="gmail_signature"><div dir="ltr"><div><div dir="ltr"><div><div dir="ltr"><div dir="ltr"><div dir="ltr"><div dir="ltr"><div dir="ltr"><div><span style="font-size:12.8px">Timothée Masquelier - <a href="mailto:timothee.masquelier@cnrs.fr" target="_blank">timothee.masquelier@cnrs.fr</a> - </span><span style="font-size:12.8px"><a href="http://cerco.ups-tlse.fr/~masquelier/" target="_blank">http://cerco.ups-tlse.fr/~masquelier/</a></span><br></div></div></div></div></div></div></div></div></div></div></div>
</div>