Sign In

Yulia Sandamirskaya

Cognitive Computing in Life Sciences, ICLS/ZHAW

Yulia Sandamirskaya

Yulia Sandamirskaya is a professor and the Head of the "Cognitive Computing" group at the Institute of Computational Life Sciences (ICLS) at the Zurich University of Applied Sciences (ZHAW) in Switzerland.1 Her LinkedIn username is yulia-sandamirskaya-0076553.

Some key points about Yulia Sandamirskaya:

  • She has over 500+ connections on LinkedIn1
  • Her research involves employing Physics-Informed Neural Network (PINN) models to predict the yield strength of materials2
  • She recently gave a thrilling and rewarding robotics lecture to over 500 elementary school students at the Kids University event5
  • She has been part of the Intel Incubation Program, which she found to be a special and amazing learning experience4
  • She has shared posts about an upcoming ICRA 2024 conference debate3 and a junior professorship position at Université Cote d'Azur that may interest young neuromorphs6

Yulia Sandamirskaya is an accomplished professor actively engaged in cutting-edge research in cognitive computing and neural networks. Her work spans both academic research and outreach to inspire the next generation of scientists.

Highlights

Jun 28 · linkedin

Those who know me and understand my research might wonder about these two beautiful papers: they are not typical.

However, at a closer look you might see the usual attractor dynamics, recurrent loops, and ‘designed’ neural architectures, even if formulated with a new maths of hyperdimentional computing and vector-symbolic architectures. No, I didn’t become a fan of large random vectors and haven’t given up metric spaces and dynamic neural fields. I see VSA representations as a sometimes convinient projection, which might make ‘the design’ feel less artificial.

We show how a scene representation can be decomposable, not entangled as with DNNs, leading to quite a few advantages, like transparency, (meaningful, i.e. functional) modularity, and fast learning. Visual odometry is one practical example. Yes, it is a very ‘prototypy’ work, yes, the examples are simple, and yes, matching hardware and software tools would help. We are on it! 🚀

Thanks a lot to Alpha Renner for pushing this work all the way to successful publications, to Paxon Frady and Fritz Sommer for collaboration and sharing their expertise and insight in the VSA/HDC world, Lazar Supic for his enthusiasm and nailing the robotic implementation, Giacomo Indiveri and Andreea Danielescu for shaping the manuscript and helping to position this work. And, of course, our institutions for supporting this quite ground-breaking and ‘blue-sky’ research: Intel Labs, Accenture, University of Berkley, ZHAW ICLS Institute of Computational Life Sciences, ZHAW Zurich University of Applied Sciences, Institute of Neuroinformatics, University of Zurich and ETH Zurich.

The papers (Springer kindly allows to read them for free 👏):

https://rdcu.be/dL57M

https://rdcu.be/dL57G

https://lnkd.in/dmifXTMV

https://lnkd.in/dwPHnpPr

Apr 2 · linkedin

This was the most thrilling and rewarding experience of the year to give a Robotics lecture to 500+ 3-6 grade kids at the Kids University @uzh. So much interest, so many good questions, smart kids: like girls boasting about their programming lessons or complaining that theirs are not good enough =o. Many thanks to RSL@ETH for the UniTree dog, which was a hit.

Related Questions

What are some of Yulia Sandamirskaya's notable research projects?
How did Yulia Sandamirskaya get involved in cognitive computing?
What is the Intel Incubation Program that Yulia Sandamirskaya participated in?
Can you provide more details about Yulia Sandamirskaya's work with Physics-Informed Neural Networks (PINN)?
What are Yulia Sandamirskaya's contributions to the field of robotics education?
Yulia Sandamirskaya
Yulia Sandamirskaya, photo 1
Yulia Sandamirskaya, photo 2
Add to my network

Location

Zürich Metropolitan Area