Makis Ntouskos

Associate Professor @UniMercatorum specializing in #AI, #ComputerVision, #DeepLearning and #Robotics; Co-founder and research director @WorldUAVHellas

[Makis Ntouskos @mdouskos]

@CSProfKGD It's more common than it should be, unfortunately.

[Angjoo Kanazawa @akanazawa]

Exciting news! MegaSAM code is out🔥 & the updated Shape of Motion results with MegaSAM are really impressive! A year ago I didn't think we could make any progress on these videos: https://t.co/rgHL4jWrsS
Huge congrats to everyone involved and the community 🎉

[Alex Dimakis @AlexGDimakis]

This paper seems very interesting: say you train an LLM to play chess using only transcripts of games of players up to 1000 elo. Is it possible that the model plays better than 1000 elo? (i.e. "transcends" the training data performance?). It seems you get something from nothing, and some information theory arguments that this should be impossible were discussed in conversations I had in the past. But this paper shows this can happen: training on 1000 elo game transcripts and getting an LLM that plays at 1500! Further the authors connect to a clean theoretical framework for why: it's ensembling weak learners, where you get "something from nothing" by averaging the independent mistakes of multiple models. The paper argued that you need enough data diversity and careful temperature sampling for the transcendence to occur. I had been thinking along the same lines but didn't think of using chess as a clean measurable way to scientifically measure this. Fantastic work that I'll read I'll more depth.

[Makis Ntouskos @mdouskos]

Our #RSLab team had a very productive field trip at #Poros, testing our #Slocum autonomous underwater #glider vehicle in the context of the @ramones_eu project.
(Article in greek)

https://t.co/1aux4Gx1zh

[Makis Ntouskos @mdouskos]

It has been a great experience presenting our research projects #iSEAu, @ramones_eu and @iToBoS_eu at the 1st Automation & Robotics Expo. Many interesting discussions with the attendees of the expo regarding the goals and the results of the projects.

[Makis Ntouskos @mdouskos]

At 14:00 GMT (15:00 CET) I am going to give a talk on "AI for Intelligent Sensing and Analysis of Underwater Environments", reserve a spot at: https://t.co/vCWl4YuRWW

@Bruneluni @AIBrunel @CBASS_Research @ntua #ArtificialIntelligence

[Dr. Asress Adimi Gikay, PhD @DrAsressGikay]

Wed, 20 Mar 2024 14:00 - 15:00 GMT: "AI for Intelligent Sensing and Analysis of Underwater Environments" by Dr Valsamis Ntouskos(https://t.co/gdtQ5pnnpi).
@CBASS_Research

[Makis Ntouskos @mdouskos]

Article about our recent expedition to Kolumbo submarine volcano near Santorini for installing novel monitoring instruments and cameras
https://t.co/nBdp5Jw5eS

[ramones_eu @ramones_eu]

Check out the @ramones_eu project as they test their prototype instruments in the hydrothermal field of Milos Island! 🌊🔬🧑‍🔬
https://t.co/DicrHKpHuE
#oceanexploration #marinebiology #hydrothermalvents #research

[Makis Ntouskos @mdouskos]

Had a great time last week at Milos island collaborating with researchers from all over Europe doing field work for #radioactivitymonitoring and #underwaterperception in the context of the EU research projects @ramones_eu (#H2020 @FutureTechEU) and #iSEAu (#H2020 @MSCActions)

[Kevin Patrick Murphy @sirbayes]

“Choosing a distribution is not a statistical question; it is a modeling decision. No statistical test can tell you whether a particular distribution is a good model for your data” - yes!

[Makis Ntouskos @mdouskos]

@GMFarinella Congratulations!

[NEANIAS Project 🇪🇺 @Neanias_eu]

NEANIAS organized the Atmospheric-workshop at the Department of Physics of the National and Kapodistrian University of Athens, with the aim of presenting the work that is active in the field of open science and in particular the atmosphere. Keep reading at https://t.co/t74RwrJn4C

[Jia-Bin Huang @jbhuang0604]

CVPR: We welcome new students!

Students: Cool! I just submitted my very first paper!

CVPR: You now also need to review papers.

Students: Not sure how to do that. Here is my attempt.

CVPR: Your reviews are too unprofessional. You are now banned for two years.

Students: ...

[Kostas Daniilidis @KostasPenn]

#CVPR2022 Motion 3: "Any reviewer who has accepted an invitation to review but violates the reviewing guidelines set forth by the conference will be prohibited from submitting any papers to CVPR for up to two
years." My vote is a resounding NO! 1/n

[Andrew Fitzgibbon @Awfidius]

So many transformer explanations want me to intuitively understand what's happening. I just want to understand what the code does. This is 25 lines of pure functional code which helped me, and might others.

https://t.co/0LstU86SpO

[Andrea Tagliasacchi @CVPR @taiyasaki]

Finally somebody run a nice ablation on all the (Ne)RF-class training acceleration papers! (InstantNGP, DVFo, Plenoxels)

This is a *must* read

[NEANIAS Project 🇪🇺 @Neanias_eu]

The workshop will start in a few minutes.
NEANIAS Underwater team will discuss and demonstrate novel, cutting-edge cloud services tackling critical aspects of underwater mapping tasks. #OpenScience
Join the Workshop at https://t.co/FYkW7VJYTM

[iToBoS Project 🇪🇺 @iToBoS_eu]

Over the last half century, clinicians and researchers have made enormous advances in detecting and treating melanoma. The common element is clinical data: images, histopathology results and information from clinicians and patients. https://t.co/kj36k1uHr1

[NEANIAS Project 🇪🇺 @Neanias_eu]

Today is #EarthDay2022 !! NEANIAS aims to offer innovative digital solutions for the underwater world, the atmosphere and space that help us to know and care for the planet. Let's have a nice and sustainable world!!

[iToBoS Project 🇪🇺 @iToBoS_eu]

The goal of the full-body scanner is to take hundreds or thousands of high-resolution pictures covering the whole body of the patient in one go and create a full avatar in which the doctors can see the details. Learn more about the iToBoS optics challenges https://t.co/ihsY6FviPd