Cornelia Fermuller

Research Scientist at UMD.
Bio-inspired machine Vision. Make robots see and understand.

[Cornelia Fermuller @CFermuller]

Last week, our project with Irina Muresanu on AI-assisted violin education was featured in the media, showing nicely the team effort between @umiacs and @TerpsMusic that has gone over the last 3+ years into this effort:
WTOP https://t.co/YNKwyiwhxt­,
WJLA https://t.co/ehVwVwEkhm

[NeuroPAC @Neuro_PAC]

Announcing the Forum on Neuromorphic Control, Tuesday, May 16, 15:00 - 16:30 ET with R. Sepulchre (@Cambridge_Uni ), M. Heemels (@TUeindhoven), and J. Cortez (@UCSDJacobs).
https://t.co/JJRCv5veie

[Cornelia Fermuller @CFermuller]

Enjoyed the article https://t.co/NmekCeDEPe by @sejnowski (@salkinstitute) on LLMs. Terry suggests that LLMs may be a mirror that reflects the intelligence of the interviewer. Later sections make interesting connections to neuroscience and suggest future research.

[Cornelia Fermuller @CFermuller]

An article in Quanta magazine https://t.co/oGLYlcmV7d describes the approach of Hyperdimensional Computing, and cites @PenttiKanerva, Bruno Olshausen (@RedWood_Neuro), @Rahimi89018256, Xun Jiao (@NovaEngineer) and myself.

[NeuroPAC @Neuro_PAC]

Announcing the first seminar in the NeuroPAC series:
Forum on Neuromorphic Navigation with panelists: Andrew Davidson (@imperialcollege)
@KostasPenn (@Penn @GRASPlab), @maththrills (@QUT) Tue, April 18, 2023, 19:00 – 20:30 GMT, 21:00 CET, 15:00 ET https://t.co/DUp4Oxh816

[UMIACS @umiacs]

A @UofMaryland expert in computer vision and world-renowned concert violinist are collaborating on the development of vAIolin—a violin teaching platform powered by artificial intelligence.

Watch the full video: https://t.co/r9ggAkIgtt

[Cornelia Fermuller @CFermuller]

In this picture, on the second road from the left, the cars seem to be moving a bit. I guess it's related to the effect of "red expanding" and the "snake illusion effect" as in Akiyoshi Kitaoka's "Rotating illusory sakura" https://t.co/Vg9gCDkH40

[Cornelia Fermuller @CFermuller]

@AkiyoshiKitaoka Hi Akiyoshi,

I wrote with Robert Pless and Yiannis Aloimonos years ago a paper on the different patterns that produce the Enigma effect (and of course, we proposed an explanation). This one is also in the paper: https://t.co/EiDeZTY0wx
website: https://t.co/jeiEsPqWjn

[Cornelia Fermuller @CFermuller]

@OlexandrBodnia @AkiyoshiKitaoka @andante_sen This demonstrates what we know from our computations. That is, estimating our own 3D motion (the motion of the person inside the moving train shooting the scene) is very challenging (and ambiguous) if the field of view is small. Great demo!

[Cornelia Fermuller @CFermuller]

@AkiyoshiKitaoka A very nice demonstration!