Profile

Cover photo
Vlado Handziski
Works at Technische Universität Berlin
Attended Technische Universität Berlin
Lives in Berlin, Germany
AboutPostsPhotosReviews

Stream

Vlado Handziski

Shared publicly  - 
 
 
PLEASE DO NOT try this at home... or at work... or anywhere. Leave it up to professionals and YouTuber ProfExtreme.

'The Professor' has used a Raspberry Pi Zero (plus other electrical bits and what-nots) to play midi files with absolutely brilliant results.

Check out their other videos for more 'Tesla coil meets Raspberry Pi' goodness including two coils playing the Airwolf theme.

(I offer no apology for when this song gets stuck in your head!)
16 comments on original post
2
Add a comment...

Vlado Handziski

Shared publicly  - 
 
 
Mythbuster 2016 - Das Nachschlagewerk (33 Seiten) ist nunmehr online verfügbar (Als Pdf Download). Martin Rotta hat sich wirklich sehr ins Zeug gelegt!
http://homepage.hispeed.ch/martin_rotta/Mythbuster-Elektroauto-Rotta.pdf
http://bit.do/mythbuster-eauto
http://tff-forum.de/viewtopic.php?f=22&t=11719&start=20
 ·  Translate
6 comments on original post
1
Add a comment...

Vlado Handziski

Shared publicly  - 
 
 
Happy to see this being discussed publicly!
'BeyondCorp' plan prefers analysis of user behaviour and device state instead
View original post
1
Add a comment...

Vlado Handziski

Shared publicly  - 
1
Add a comment...

Vlado Handziski

Shared publicly  - 
 
 
The fascinating impact on Korean society of AlphaGo's 4-1 win over Lee Sedol. President Pak Geun-Hye: “Above all, Korean society is ironically lucky, that thanks to the ‘AlphaGo shock’, we have learned the importance of AI before it is too late.” http://www.nature.com/news/south-korea-trumpets-860-million-ai-fund-after-alphago-shock-1.19595

It begins. http://xkcd.com/1656/
Historic win by Google DeepMind's Go-playing program has South Korean government playing catch-up on artificial intelligence.
1 comment on original post
1
Add a comment...

Vlado Handziski

Shared publicly  - 
 
 
PlaNet:Taking photo geolocation to the next level. This paper raises the bar for photo-based localization. It departs from the retrieval-based paradigm used so far and proposes a smart way of dividing the earth surface in adaptive cells depending on the photo/population density of the area. This leads to ~26k locations which are further use as classes to train a deep network. Training on such task should be difficult and lengthy as the classes are numerous and images are noisy. The authors end up using 125M images for training over 2.5 months on 200 CPU cores. The results are impressive and the method can be further improved with an LSTM module trained over albums of images to get a sense of memory and co-localization of images from the same album.
I do hope the model will be made public by the authors as it could open up a lot of related applications and ideas.

Abstract: Is it possible to build a system to determine the location where a photo was taken using just its pixels? In general, the problem seems exceptionally difficult: it is trivial to construct situations where no location can be inferred. Yet images often contain informative cues such as ...
View original post
1
Add a comment...

Vlado Handziski

Shared publicly  - 
 
 
There's more than one way to draw an ellipse.
3 comments on original post
1
Add a comment...

Vlado Handziski

Shared publicly  - 
 
 
TL;DR Russian photographer used neural networks to find people he saw on the subway on social networks. And he compared those real life experiences with how those people present themselves. Fascinating. Frightening. https://birdinflight.com/ru/vdohnovenie/fotoproect/06042016-face-big-data.html
 ·  Translate
Фотограф Егор Цветков с помощью общедоступных программ ищет и находит в социальных сетях людей, которых снимает в метро.
6 comments on original post
2
Add a comment...

Vlado Handziski

Shared publicly  - 
 
 
Also - this is fun. :)
Tinker with a real neural network right here in your browser.
View original post
2
1
Add a comment...

Vlado Handziski

Shared publicly  - 
 
 
#Nodejs is quickly becoming the go-to platform for full stack development. Which is why we’re excited to join the Node.js Foundation—and to help the framework keep growing. Learn more about how we’ll be working together: https://goo.gl/Uo4xOF
1 comment on original post
1
Add a comment...

Vlado Handziski

Shared publicly  - 
 
 
Interesting article on smart devices, concluding with "increased complexity has both advantages and pitfalls ... Essentially, full-on computing devices with multiple communications channels, upgradeable firmware, linked cloud services, and apps. Any of these bits can fail ...The upsides of all this complexity are incredible features, such as the ability to know if your home in Minnesota is suffering from a carbon monoxide leak even while you're sitting on the beach in Florida ... Welcome to the Internet of Things... and its tradeoffs."
View original post
1
Add a comment...

Vlado Handziski

Shared publicly  - 
1
Add a comment...
Story
Tagline
Researcher, husband, father of two.
Introduction
I am member of the Telecommunication Networks Group at the Technische Universität Berlin working in the area of networked embedded systems. 
Education
  • Technische Universität Berlin
    Dr.-Ing. (summa cum laude), Electical Engineering and Computer Science, 2002 - 2010
  • Ss. Cyril and Methodius University
    MSc., Electical Engineering and Computer Science, 1998 - 2002
  • Ss. Cyril and Methodius University
    Dipl.-Ing., Electical Engineering and Computer Science, 1993 - 1998
Basic Information
Gender
Male
Relationship
Married
Other names
Владо Ханџиски
Work
Occupation
Postdoc/Interim Professor at TU Berlin and TU Dresden
Employment
  • Technische Universität Berlin
    Postdoc/Assistant Professor, 2011 - present
    Telecommunication Networks Group www.tkn.tu-berlin.de
  • Technische Universität Dresden
    Interim Professor, 2014 - 2014
    Chair for Embedded Systems http://www.es.inf.tu-dresden.de/
  • Technische Universität Berlin
    Research Associate, 2002 - 2010
  • Ss. Cyril and Methodius University
    Teaching and Research Assistant, 2000 - 2002
Places
Map of the places this user has livedMap of the places this user has livedMap of the places this user has lived
Currently
Berlin, Germany
Previously
Washington DC, USA - Skopje, Macedonia
Contact Information
Work
Phone
+49 30 314 238 31
Email
Address
Sekr. FT5, Einsteinufer 25, 10587 Berlin, Germany