Profile

Cover photo
Tobias Weyand
Works at Google
Attended RWTH Aachen
Lives in Venice
157 followers|885,798 views
AboutPostsPhotosVideosReviews

Stream

Tobias Weyand

Shared publicly  - 
 
 
Turning black-and-white photographs into colour photographs with a bit of deep learning.
I really enjoyed this one - it is cool and nicely written. The author is the original designer of node.js. As far as I know he doesn't have a background in machine learning. I wonder if magic will soon be in every good programmer's toolbox?
2 comments on original post
2
1
Konstantinos Rematas's profile photo
Add a comment...

Tobias Weyand

Shared publicly  - 
 
Some good laughs in here!
 
http://clickotron.com/
Articles generated by a recurrent neural network trained on Buzzfeed:
"How The World's Most Extreme Baby Moms Lost Weight"
"David Beckham Looking For Them In "Lady Box""
"This Guy Thinks His Cat Was Drunk For His Five Years, He Gets A Sex Assault At A Home"
"What To Use When You're A Great Person"
View original post
1
Add a comment...

Tobias Weyand

Shared publicly  - 
 
 
All of these images were computer generated!

For the last few weeks, Googlers have been obsessed with an internal visualization tool that Alexander Mordvintsev in our Zurich office created to help us visually understand some of the things happening inside our deep neural networks for computer vision.  The tool essentially starts with an image, runs the model forwards and backwards, and then makes adjustments to the starting image in weird and magnificent ways.  

In the same way that when you are staring at clouds, and you can convince yourself that some part of the cloud looks like a head, maybe with some ears, and then your mind starts to reinforce that opinion, by seeing even more parts that fit that story ("wow, now I even see arms and a leg!"), the optimization process works in a similar manner, reinforcing what it thinks it is seeing.  Since the model is very deep, we can tap into it at various levels and get all kinds of remarkable effects.

Alexander, +Christopher Olah, and Mike Tyka wrote up a very nice blog post describing how this works:

http://googleresearch.blogspot.com/2015/06/inceptionism-going-deeper-into-neural.html

There's also a bigger album of more of these pictures linked from the blog post:

https://goo.gl/photos/fFcivHZ2CDhqCkZdA

I just picked a few of my favorites here.
25 comments on original post
3
1
Konstantinos Rematas's profile photo
Add a comment...

Tobias Weyand

Shared publicly  - 
 
There's a new podcast on machine learning called Talking Machines. The first episode sounds promising. http://www.thetalkingmachines.com/blog/
2
Add a comment...

Tobias Weyand

Shared publicly  - 
 
I have a little favor to ask: For an upcoming paper, I'm creating a new public ground truth for evaluating landmark recognition systems. If you have a bit of time, it would be awesome if you could help me out with creating the annotations. It's super-easy and fun! Here's how it works:

https://docs.google.com/presentation/d/1Af6EEXdf6i6Fu0kVUxPK-y94zjzexbILEV6jQTWPJF4/pub?start=false&loop=false&delayms=3000

The annotation tool is here:

http://recog.vision.rwth-aachen.de:31415/

Every little contribution counts! Thank you already for your help and please spread the word!
4
Rebecca Hosang's profile photo
 
Ich bin leider zu dumm dazu am Handy...
 ·  Translate
Add a comment...
Have him in circles
157 people
René Bohne's profile photo
Tobias Eric Perez Alvarez's profile photo
Rodrigo Benenson's profile photo
Hans-Christian Ebke's profile photo
Friends of Nature's profile photo
Javier Marin's profile photo
Eduard Feicho's profile photo
Glailton Luis's profile photo
Li Ming's profile photo

Tobias Weyand

Shared publicly  - 
 
 
The next generation of applications won’t just store images. They’ll actually understand them thanks to models like the one we’re releasing today. Our Cloud Vision API uses machine learning to classify images into thousands of categories, detect objects (and facial features), and even identify language. We’re excited to see what you build with it: http://goo.gl/RfEFv0
7 comments on original post
1
Add a comment...

Tobias Weyand

Shared publicly  - 
 
 
Machine learning has had a huge impact on computer science, making it possible explore new frontiers of research and to develop amazingly useful products that millions of people use every day. Our machine learning system, TensorFlow, is used for everything from speech recognition in the Google app, to Smart Reply in Inbox, to search in Google Photos. 

And now you can use it too.

Today, we’re proud to announce the open source release of TensorFlow - our second-generation machine learning system that is a general, flexible, portable, easy-to-use, and completely open source software library for numerical computation using data flow graphs. Head over to the Google Research blog to learn more.
7 comments on original post
2
Add a comment...

Tobias Weyand

Shared publicly  - 
 
 
Amazing results on novel view synthesis using deep learning... This is the paper http://arxiv.org/abs/1506.06825

(Via +Kosta Derpanis​)
5 comments on original post
3
Add a comment...

Tobias Weyand

Shared publicly  - 
 
New Paper:

Fixing WTFs: Detecting Image Matches caused by Watermarks, Timestamps, and Frames in Internet Photos

http://www.vision.rwth-aachen.de/projects/fixing-wtfs
Detecting Image Matches caused by Watermarks, Timestamps, and Frames in Internet Photos. examples.png. Summary. A common problem of computer vision applications that are based on Internet photos are false-positive matches caused by Watermarks, Timestamps, and Frames (WTFs) superimposed on the ...
6
Torsten Sattler's profile photoIaroslava Grinchenko (Grinning)'s profile photo
2 comments
 
WTF. Cool name :D
Add a comment...

Tobias Weyand

Shared publicly  - 
 
 
Project Tango wouldn't be possible without the decades of research from around the world that has gone into computer vision and robotics. We thought it appropriate to highlight some of this great work and show how it has direct impact on the future of mobile devices with a more human-like sense of space and movement.

These are just a small portion of the many great engineers whose work has gone into Project Tango. But it's some amazing stuff and I hope everyone enjoys this look behind the scenes of some phenomenal research labs.
1
Add a comment...

Tobias Weyand

Shared publicly  - 
 
Put together a little tool to create 3D point clouds from the depth maps that Google Camera's new Lens Blur feature generates: https://github.com/0xtob/gcam2ply
12
Add a comment...
People
Have him in circles
157 people
René Bohne's profile photo
Tobias Eric Perez Alvarez's profile photo
Rodrigo Benenson's profile photo
Hans-Christian Ebke's profile photo
Friends of Nature's profile photo
Javier Marin's profile photo
Eduard Feicho's profile photo
Glailton Luis's profile photo
Li Ming's profile photo
Education
  • RWTH Aachen
    2015
Basic Information
Gender
Male
Work
Employment
  • Google
    present
Places
Map of the places this user has livedMap of the places this user has livedMap of the places this user has lived
Currently
Venice
Previously
Aachen - Paris
Links
Sehr nett und super lecker! Unbedingt das Cornelius-Bräu probieren!
Public - a year ago
reviewed a year ago
1 review
Map
Map
Map