Profile

Cover photo
jimun gimm
68,395 views
AboutPostsPhotosVideos+1's

Stream

jimun gimm

Shared publicly  - 
 
 
For real or not for real, I'm curious about how a Flyboard works (or can work).
View original post
1
Add a comment...

jimun gimm

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.
24 comments on original post
1
Add a comment...

jimun gimm

Shared publicly  - 
 
7 Exciting Financial Apps disrupting the Industry: The financial sector is no exception to technological disruption, with these apps below revolutionising everything from personal finance to stock trading and accounting.
1
Add a comment...

jimun gimm

Shared publicly  - 
 
36 world-leading PR, content marketing, social media and SEO experts reveal how karma is king for a new breed of communicators.
1
Add a comment...

jimun gimm

Shared publicly  - 
 
Introducing Hands Free payments, from Google.
1
Add a comment...

jimun gimm

Shared publicly  - 
 
Yonatan Zunger originally shared to Today I Learned::
 
One of the biggest challenges in information retrieval (the branch of computer science that includes search and content recommendation) is how to find good content which humans haven't already found. To date, the most reliable signals have been other human judgments: for example, PageRank is a measure of how "good" a site is based on links people have made to that site (with the challenge being how to separate "meaningful" and trustworthy links from the rest), and collaborative filtering is based on what other users have chosen (with the challenge being how to find users with similar enough taste to be relevant).

The challenge is that, when new material shows up on the scene, you don't yet have any human interactions -- and quite often, good material, things people would love, simply goes unnoticed and never builds up the interaction signals which help. To detect quality in these things requires understanding the content itself, and the aspects of it which matter to people.

There are several hard aspects to this. One is simply understanding the content at the right granularity: "the color of the top-left pixel" or "the frequency of the word 'whenever'" are too fine-grained to give us a hint about whether people will like something, so we need to be able to group the content into more meaningful structures. For images, that might be "an image of a face in 3/4-profile," a certain color balance or contrast, a perspective or a cropping, and advances in image recognition in the past few years have (finally) made it possible to reliably identify such features. For text, it's much harder: there isn't yet even a clear idea of what features both could be measured about text and determine people's tastes. (How do you measure "intellectually meaty" or "hinting at scandal?")

This paper has used the recent advances in image processing, together with recent advances in AI in general, to get a sense of which pictures people will like. It started by taking several thousand images, and having them rated by humans for quality; that was used as "ground truth." Then, those thousands of images are analyzed into meaningful features, and a neural network is trained to find patterns of image features which predict human taste.

This is what neural networks, and other kinds of "supervised" machine learning systems, do in general: they take as inputs a bunch of signals, and combine them using a large number of parameters -- the "weights" -- to produce predictions of some values that you want to measure. The weights are set by taking a large number of test examples ("golden data" or "ground truth") with known values of both the signals and the test values; weights are chosen ("trained") to maximize the quality of the system's predictions for this data. To make sure that the training doesn't just teach it to recognize those specific examples, the golden data is randomly split into two groups; one is used for training, and then it's tested against the other group to make sure that the predictions with the trained weights are good. If they are, then you have a model which can predict -- given any set of measured signals -- the truth values.

In this case, the signals are these features of the image, measured by a second machine learning system; the quantity being predicted is whether people will like it. Because these are all "content-based signals" -- that is, they're based on the contents of the image, and not on people's responses to it -- the resulting model can be applied to any image. 

The team then applied this model to a set of 9 million images from Flickr with fewer than five "favorites." They tested the quality of its picks by having human raters compare that result set with the set of popular images on Flickr; the result was excellent, with its "hidden gems" scoring statistically the same as the most popular images on the site.

I would expect a lot more work on related techniques over the next few years, and for this to have a significant impact on the way that content recommendation is done. The main upshot will be that more little-known works get the spotlight they deserve -- something critical, as more and more people are creating things of value that they want the world to see. 

h/t +Wayne Radinsky and +Daniel Estrada
Beautiful images are not always popular ones, which is where the CrowdBeauty algorithm can help, say computer scientists.
138 comments on original post
1
Add a comment...

jimun gimm

Shared publicly  - 
 
Human connection is the most compelling of things about 'social'. And in the early days, Google+ connected people in new ways; with 'Hangouts' the doors of the
1
Bill G's profile photo
Bill G
 
Well all that mumbo jumbo, but "collections" are still a detour to dead ends.
Add a comment...

jimun gimm

Shared publicly  - 
 
It seems like an eternity ago that we started our weekly list of the best new tools for startups. But really it has only been about two months since we began this journey to find tools that we thought could help out our fellow entrepreneurs! After accumulating more than a hundred unique tools, we thought that
1
Add a comment...

jimun gimm

Shared publicly  - 
 
Keyword research is the foundation to getting to #1 in Google and getting the TRAFFIC you need to scale your business through SEO. Here's how to do it...
1
Add a comment...

jimun gimm

Shared publicly  - 
 
 
Photographer Sasi Smith Captures the Mesmerizing Spirit of Adorably Wise Owls #photography   #owl   #nature   #SasiSmith  

See more: http://www.mymodernmet.com/profiles/blogs/sasi-smith-adorable-owl-photography
3 comments on original post
1
Add a comment...
People
Collections jimun is following
View all
Basic Information
Gender
Male
Story
Introduction
I might be your friend.
Work
Occupation
Industrial Technology Security Consultant/Web Designer
jimun gimm's +1's are the things they like, agree with, or want to recommend.
The Future of Google+ - Martin Shervington
www.martinshervington.com

The Future of Google+ is based around Collections and Communities. This Stream will help you connect with great people and ideas, broadening

조선시대 우리 옷의 멋과 유행 - Google Cultural Institute
www.google.com

Google Cultural Institute는 숨결을 불어넣는 스토리를 가미한 수많은 역사적 유물을 여러 파트너로부터 협찬받아 가상 박물관에 한 데 모으고 있습니다.

Revealed at Last: Magic Leap's Vision for Augmented Reality, in 32 Paten...
www.wired.com

Here's why Google invested $542 million in an secretive Florida startup.

Google+ Your Business
plus.google.com

Official page for businesses using Google to connect with people

영상으로 돌아보는 2014 유튜브: 만들고, 즐기고, 공유한 모든 순간을 기념합니다
youtubekrblog.blogspot.com

올해도 유튜브에는 전세계적으로 회자되며 대중문화 역사에 길이 남을 다양한 볼거리들로 가득했습니다. 폴란드에서 등장한 대형 거미 개에 혼비백산 하거나, 여성 혼자 뉴욕 거리를 10시간 동안 활보할 때 사람들의 반응을 관찰한 몰래 카메라 영상, 처음 보는

[인터뷰] 한국 문화를 사랑하는 유튜버들의 특별한 오프라인 프로젝트!
youtubekrblog.blogspot.com

구글 APAC 블로그에서는 인터넷을 통해 창작과 만남, 그리고 성장을 이루며 아시아 지역에서 주목 받고 있는 유튜버들의 이야기를 시리즈로 공유하고 있습니다. 금주에는 한국의 대표 유튜브 창작자로 꼽히는 두 팀의 특별한 ‘오프라인 프로젝트’를 소개하려고

99 Sites That Every Professional Should Know About and Use
www.themuse.com

The internet is a big place, but we've found the best of it.

Video Publishing by Robin Good | ZEEF
video-publishing.zeef.com

The best tools and services to capture, edit, publish and distribute video online.

What exactly is the Internet of Things? A brief primer- Postscapes
postscapes.com

An overview handbook looking into what the Internet of Things is and how it is coming into being

iot.eclipse.org - IoT development made simple
iot.eclipse.org

iot.eclipse.org is where you can learn about the technologies developed at Eclipse to make Internet of Things (IoT) development simpler. The

Bunkr is a collaborative online presentation tool
www.bunkr.me

bunkr, presentation, tool, cloud, slides, productivity, best, great slides, powerpoint killer, creativity

Circle Master
plus.google.com

Share circles created by the Circle Master to grow on Google+. #CircleMaster

A Facebook Life | TechCrunch
techcrunch.com

On Facebook, life begins at conception. "We're expecting!", your parents post. You don't have fingers but you're already accruing likes. A s

Content Curation by Lucian Duma | ZEEF
contentcuration.zeef.com

The Internet Curator’s role is to seek on the web information related to a specic domain,fillter them, select them,organise,share,present th

A collection of hand selected articles by Jimun Kim from SocialMediaDesign
bit.ly

Now in Beta! Paper.li multi-share - share your favorite articles, photos and videos from one paper in 1 go.

Steve Jobs talks about branding | StockLogos.com
likehack.com

Find and purchase an exclusive, unique and professional logo now. Customization for your name is included in the price. Read more

Life Advice From 18 of the Wealthiest People in History (Interactive Gra...
www.entrepreneur.com

Wise words from Bill Gates, Mark Zuckerberg, Michael Dell, Elon Musk, Andrew Carnegie and other legendary leaders on success, attention, mot

Les médias sociaux : Guide d'utilisation
www.scoop.it

Les médias sociaux : Guide d'utilisation Un petit guide rapide sous la forme d'une infographie Via Tolokonnikoff - Seratoo

Intersection of Social Media and Storytelling
www.ishmaelscorner.com

The Art Of Storytelling In Business Communications And Public Relations