Image Search: Getting Better and Better
May 15, 2015
Image search means having software which can figure out from a digital photo that a cow is a cow. In more complex photos, the software identifies what it can. I recall one demonstration which recognized me as a 20 year old criminal. Close but no cigar.
I received an email from a former clandestine professional. The link provided informed me that Baidu was better at image recognition than the Google. The alleged error rate is 4.58 percent. I love the two decimal accuracy.
Not to be outdone, WolframAlpha is in the image recognition game as well. Navigate to “Wolfram Alpha Image Identification Identifies Steven Wolfram as Podium.” The write up points out:
Speaking of which, a picture of Steven Wolfram returned the answer ‘podium’. So no recognition for the creator. Unfortunately, it couldn’t identify a map of France at all and just came back with a big question mark. Sorry, France.
You can try the system at this page.
I uploaded the image of the cover of my new CyberOSINT study. The system returned this result:
My book cover is a a piece of electronic equipment that mixes two or more input signals to give a single output signal.
I did not know that. I thought it was a book cover with a blue hand.
Stephen E Arnold, May 15, 2015
Visual Browsing: A New, Next, Big Thing. Maybe.
April 29, 2015
The visual browsing bandwagon is rolling along. The sponsored content Guardian in the UK published “Visual Browsing: There’s a Critical Gap between How We See and How We Search.” The write up, which seems to be supported by SAP, states:
What we need is a visual browser for the world around us – a way of pointing at things which inspire thoughts and questions, giving us a rich, engaging means to find out what we don’t know, and those things we didn’t know how to search for using mere words.
Right, words. The challenge according to Blippar, the outfit connected with this visual search, essay points out:
Visual browsing sits at the heart of discovery in the internet of everything. It has the potential to bring the world to life around us, adding a story to every thing we see and the ability to sate our curiosity in every moment. Visual browsing is the most ‘native’ search engine there is, being based on context alone, driven by visual cues, location, time of day and the interests of the user, and not biased or limited by the understanding or vocabulary of the user.This will give us the ability to satisfy our curiosity more of the time – to visually search for the answers to the questions that intrigue us every day; to truly take search into the realm of ‘discovery’. We’re the most curious of species on the planet – it’s what’s got us to where we are today. The next generation of search must reflect this.
Blippar allows a person to take a picture using a mobile phone and then having the picture generate results.
If you want to see examples of visual browsing, point your browser to Qwant.com. This is the French Web search system owned in part by Axil Springer. For an example of a browser that itself incorporates visual browsing, download a copy of Vivaldi.
A picture, according to my somewhat addled great grandmother who wrote poetry with curse words as a metaphorical trope, is worth a thousand words. Here’s Qwant’s results for the query semantic search:
Visual browsing is one component of a next generation information access system, just not a main component. Clutter is not useful when certain types of information is required under difficult conditions such as a flash crash or someone is lobbing ordinance in your direction.
I am trying to figure out the SAP and Blippar connection. Will my mobile phone snap of the SAP logo help? I think not.
Stephen E Arnold, April 29, 2016
Informed Millennials
April 15, 2015
With the fall of traditional newspapers and aging TV News audiences, just where are today’s 20- and young 30- somethings turning for news coverage? Science 2.0 tells us “How Millennials Get News,” reporting on a recent survey from the American Press Institute and the Associated Press-NORC Center for Public Affairs Research. The joint effort comes from a collaboration arrangement the organizations call the Media Insight Project. Conducted at the beginning of 2015, the survey asked Millennials about their news-consumption habits. The article tells us:
“People ages 18-34 consume news and information in strikingly different ways than did previous generations, they keep up with ‘traditional’ news as well as stories that connect them to hobbies, culture, jobs, and entertainment, they just do it in ways that corporations can’t figure out how to monetize well….
“‘For many Millennials, news is part of their social flow, with most seeing it as an enjoyable or entertaining experience,’ said Trevor Tompson, director of the AP-NORC Center. ‘It is possible that consuming news at specific times of the day for defined periods will soon be a thing of the past given that news is now woven into many Millennials’ connected lives.’”
Soon? Even many of us Gen Xers and (a few intrepid Baby Boomers) now take our news in small doses at varying hours. The survey also found that most respondents look at the news at least once a day, and many several times per day. Also, contrary to warnings from worrywarts (yes, including me), personalized news feeds may not be creating a confirmation-bias crisis, after all. Most of these Millennials insist their social-media feeds are well balanced; the write-up explains:
“70 percent of Millennials say that their social media feeds are comprised of a diverse mix of viewpoints evenly mixed between those similar to and different from their own. An additional 16 percent say their feeds contain mostly viewpoints different from their own. And nearly three-quarters of those exposed to different views (73 percent) report they investigate others’ opinions at least some of the time–with a quarter saying they do it always or often.”
Well, that’s encouraging. Another finding might surprise some of us: Though a vast 90 percent of Millennials have smart phones, only half report being online most of all of the day. See the article for more, or navigate to the report itself; the study’s methodology is detailed at the end of the report.
Cynthia Murrell, April 15, 2015
Stephen E Arnold, Publisher of CyberOSINT at www.xenky.com
Search and Identify a YouTube or Vimeo Tune
April 12, 2015
Need to identify a song used in a YouTube video? “Name That tune on Any YouTube Video with MooMa.sh” explains that now you can perform this search and retrieval task. Navigate to http://www.mooma.sh/. Paste a YouTube, Vimeo, or Dailymotion link into the search box and Moo1. That’s the service’s name for search, not mine. There is a video explaining how the service works and a Freshman Comp 101 write up that explains how. I use Samba Pump, for which I paid a fee. MooMa.sh reported:
Stephen E Arnold, April 12, 2015
Digital Imaging: An IDC Estimate Misses the Mark
February 6, 2015
Digital imaging is getting bigger. I came across this interesting factoid in LensVid:
Predicting the future of the camera market proved challenging in the past – IDC (the American market research, analysis and advisory firm) failed to predict what will happen to the mirrorless camera market. In 2012 they concluded that in 2014 we will see no less than 13 million mirrorless cameras sold worldwide. Only 3 million mirrorless cameras were actually sold…
For the full run down on digital photography in 2014, navigate to “LensVid Exclusive: What Happened to the Photography Industry in 2014?”
Next time someone tosses mid tier consulting firm predictions your way, perhaps you should let them pass on by.
Stephen E Arnold, February 6, 2015
Google: More Ad Pain Coming
December 10, 2014
I am not sure if the information in “Facebook Video Is Driving YouTube Off Facebook” is spot on. Counts of user behavior without the actual log files are subject to interpretation. But the main point is darned suggestive. Facebook video may be cutting into uploads to YouTube.com. Now the Googlers are trying to make YouTube into a bigger money spinner. If Facebook pushes into video, advertisers are going to want to put their messages in front of Facebook viewers of hot videos. Bad news for Google.
The passage from the article I noted was:
It is evidence of a dramatic shift in power: Until recently Facebook was not even considered a destination for video. Page owners simply shared their YouTube videos on Facebook, and that was that.
My view is that Google struggles to convert social into a service that can compete with Facebook. If Facebook figures out how to play nice with China, the GOOG has a yellow alert flashing. Is the answer in “How Google Works”?
Stephen E Arnold, December 10, 2014
UK Paintings Catalog: When Every Does Not Mean Every
December 2, 2014
I love headlines like “Every Painting in the UK at Your Fingertips.” The idea is that “images and details of every painting (in tempera or acrylic) in public ownership through the United Kingdom.” Well, obviously the “every” is not every painting. There is an 86 volume set which presumably presents the images and metadata. The digital images are available at Your Paintings. There is a search box and a number of other options. I ran a query for Patrick Heron, an artist whose work I find interesting. There are some of his pictures in the Tate, and he was born . Here’s what I found:
Pretty thin. The Patrick Heron entry for the St Ives School offers a bit more information.
I am not sure if the BBC index is incomplete. It appears that posting information or links to other UK online sources is not part of the project. Also, the presentation of different search boxes on the BBC site does not make accessing the Your Paintings information easier.
The enthusiasm of the newspaper is admirable. I expect/hope that the service will improve its usability and completeness in the months ahead. The BBC is, as one of my British acquaintences with an Oxford education used to say, performant.”
Stephen E Arnold, December 2, 20141
Sir Thomas Bayes Does Art. Versatile Guy.
September 17, 2014
Navigate to FindMeLike. Click on “Try this demo.” You will have access to a Bayesian-centric visual search tool. The idea is that you click on an image you like. The system then locates similar images.
The click narrows the result set. Each poster is available for sale. But I could not figure out how to move to the shopping cart.
How well does a Bayesian-centric system work? Try and use the comments section of this blog to share you opinion.
Stephen E Arnold, September 17, 2014
Machine Vision Solved (Almost)
September 11, 2014
I read “The Revolutionary Technique That Quietly Changed Machine Vision Forever.” The main idea is that having software figure out what an image “is” has become a slam dunk. Well, most of the time.
The write up from the tech cheerleaders at Technology Review says, “Machines are now almost as good as human at object recognition.”
A couple of niggling points. There is that phrase “almost as good”. Then there is the phrase “object recognition.”
Read the write up and then answer these questions:
- Is the method ready to analyze imagery fed by a drone to a warfighter during a live fire engagement?
- Is the system able to classify a weapon in a manner meaningful to field commander?
- Can the system discern a cancerous tissue from a non cancerous tissue with an image output from a medical imaging system?
- Does the method recognize objects in a image like the one shown below?
Image by Stephen E Arnold, 2013
If you pass this query to Google’s image recognition system, you get street scenes, not a person watching activities through an area cordoned off by government workers.
Google thinks the surveillance image is just like the scenes shown above. Note Google does not include observers or the all important police tape.
The write up states:
In other words, it is not going to be long before machines significantly outperform humans in image recognition tasks. The best machine vision algorithms still struggle with objects that are small or thin such as a small ant on a stem of a flower or a person holding a quill in their hand. They also have trouble with images that have been distorted with filters, an increasingly common phenomenon with modern digital cameras.
This stuff works in science fiction stories, however. Lab progress is not real world application progress.
Stephen E Arnold, September 11, 2014
Instagram Search from Picturegr.am
September 3, 2014
If you are interested in searching Instagram images, navigate to www.picturegr.am.
The site says:
Picturegr.am is a new Instagram search engine and web viewer. Featuring millions of pictures, users, likes and comments, Picturegr.am is your go-to source when you want to browse Instagram on computer or desktop. Picturegr.am works on both PC and Mac.
A user can query Picturegr.am by hashtags or user name. Instagram users assign hashtags and their handles. As a result, a query for “visualization” returns images and terms; for example, on September 2, 2014:
A more popular hashtag like “chicagobears” returns images more in line with non specialist content; for example:
Interesting but filtering and limits on access to user content may trouble some.
Stephen E Arnold, September 3, 2014
-
- Subscribe to Beyond Search
Feature archive
News archive
-
