Images, Art, Facial Recognition, More: Sunday Morning Buzz, May 17th, 2015


Wolfram|Alpha has launched an image identification tool. “Now I’m excited to be able to say that we’ve reached a milestone: there’s finally a function called ImageIdentify built into the Wolfram Language that lets you ask, “What is this a picture of?”—and get an answer. And today we’re launching the Wolfram Language Image Identification Project on the web to let anyone easily take any picture (drag it from a web page, snap it on your phone, or load it from a file) and see what ImageIdentify thinks it is…” Warning: you can play with this for hours. I uploaded an image of one of my cats and it got it spookily correct (“Calico cat”) but then I uploaded a picture of a person and it misidentified him as a fire extinguisher. It seems to do best with images without lots of details.

Amit Agarwal, who has no need to prove how brilliant he is but keeps doing it anyway, has created a tool to send bulk personalized Tweets and DMs.


May be useful depending on your research needs: a roundup of 60 facial recognition databases.

Interesting! Using a ‘bot to help people discover art. “Artbot, developed by Desi Gonzalez and Liam Andrew in the HyperStudio research group of Comparative Media Studies/Writing (CMS/W), is a mobile website app that mines both user preferences and event tags to provide serendipitous connections to the local art scene…. Artbot enables users to select their interests from a list that ranges from medieval art to surrealism and from ancient history to photography. At the same time, the app scrapes data from museum websites to find artists, movements, and themes that link events to each other in various ways. Artbot then cross-references the data collected to generate event recommendations.”


Chromecast has gotten some updates. “Ever since Google launched the Chromecast in July 2013, the company has been steadily updating the HDMI dongle with new capabilities and features. Today, the company has announced six new apps for its $35 streaming media stick: CBS All Access, HGTV, FOX Now, FXNOW, Pluto TV, and Haystack.”

Libraries and Archives of Canada has put more WWI service files online. “As of today, 155,110 of 640,000 service files are available online…”


YouTube “How To” video searches are way up in 2015. “People trying to figure out how to accomplish a home improvement project, fix their hair or prepare a recipe have helped grow YouTube’s ‘how-to’ searches by 70 percent year-over-year.”

More YouTube: what’s YouTube’s most-watched game? Why, it’s Minecraft. “Think about that for a minute. YouTube’s list of the top 10 biggest games on the site, based on a decade’s worth of viewing hours, features long-running game franchises like Call of Duty and Grand Theft Auto. But it’s six-year-old Minecraft that comes out on top.”

From Search Engine Land: How Google made it a little harder to reach from outside the US. “Last fall, things were quietly changed. Instead of that link always being at the bottom of country-specific versions, it was altered to appear only the very first time someone tried to reach and got redirected to their country-specific version. On subsequent attempts, it would not be shown.”

There are concerns going around about a phantom Google update. “HubPages, a collection of more than 870,000 miniblogs covering everything from the ‘History of advertising’ to “How to identify venomous house spiders,” saw its Google search traffic plunge 22 percent on May 3 from the prior week. Of the company’s 100 top pages, 68 lost visitors over that stretch.”

HathiTrust, in its blog, has a post about quality and OCR issues. “For the digital content we ingest, HathiTrust has established specifications related to image formats, resolution, color space, and other characteristics. Rigorous validation ensures that these specifications are met. The methods of production or processing of digitized items may leave fingerprints of some sort, however. These may be benign, such as the presence of digitization color targets, added coversheets, book cradles, or a characteristic coloration of pages, which do not generally interfere with the display or understanding of the original object and its content. They may also be more serious, including mis-colorations of pages, human fingers in the images, systemic cropping, warping, or bolded or light text—problems that do interfere with legibility or clarity of the image.”

The Digital Public Library of America (DPLA) is now on Pinterest. Good morning, Internet…

I love your comments, I love your site suggestions, and I love you. Feel free to comment on the blog, or @ResearchBuzz on Twitter. Thanks!

IE, Heartbleed, Denmark, More: Monday Buzz, April 28, 2014

This is definitely the headline news of the morning: there is a zero-day security issue in every version of Internet Explorer. Every one. Including the ones in XP that won’t get patched. “Attacks taking advantage of the vulnerability are largely targeting IE versions 9, 10, and 11 in something called a ‘use after free’s attack. Essentially, the attack corrupts data as soon as memory has been released, most likely after users have been lured to phony websites.” If you are using IE (and I hope you aren’t) your best bet is to switch to some other browser until Microsoft issues a patch.

A little outside the normal ResearchBuzz stuff but it showed up in my Google Alerts yesterday morning and I really liked it. Rhode Island School of Design professor Clara Lieu has been doing an “Ask the Art Professor” column for a year now, and in this blog post summarizes all the available columns. There are about a hundred of them, and while some of them are very specific to art (“How Can I Learn to Draw Noses?”), many of them were relevant to the creative process in general. Professor Lieu’s column continues and is now available at the Huffington Post.

I love me some screenshot/screencast tools. Lifehacker has a writeup on one called TinyTake. Windows only, alas.

Wisconsin fishermen have a new resource for finding particular fish species. A new tool lets users search for 160 species by county or habitat types.

I’m still following Heartbleed. If you’re like me and are taking both a practical and nerdy interest in it, you’ll like this article from Rubin Xu on how he stole a server’s private key using Heartbleed.

More security: apparently a person figured out how to DDOS a site using Facebook Notes, but Facebook isn’t going to give them a bug bounty. (“…the conclusion is that there’s no real way to us fix this that would stop ‘attacks’ against small consumer grade sites without also significantly degrading the overall functionality. Unfortunately, so-called ‘won’t fix’ items aren’t eligible under the bug bounty program, so there won’t be a reward for this issue.”) The other large post on this very new blog is an overview of how a site can be DDOS’d using Google Spreadsheets, which they also won’t be getting a bug bounty for….

Lifehacker has a handy tip for finding all those forgotten accounts with a simple e-mail search. Using specific vocabularies in search is very, very handy.

BetaList has a link to GMail tool Sortd, which sounds almost too good to be true. “Sortd is a smart workspace for Gmail that lets you manage your work, tasks and email all in one place (right inside Gmail). Drag important emails out of your Inbox onto a set of personalized priority lists, where you can see a birds-eye view of everything you have on the go.”

From the design end of things: 7 Things I Wish Every Search Box Did. “Great search experience is all about speed and relevance. You want to provide the right result for minimum effort. Your product needs a search engine that thinks like your users, and one that understands from a few letters exactly what is being searched for. How do you do that? Here’s 7 ways.”

WOW: the entire country of Denmark has been replicated in Minecraft.

Joyce Valenza takes a quick look at Vellum, a New York Times experiment for content discovery on Twitter. If I didn’t have Nuzzel I’d be all over this. As it is, it’s interesting.

Ever wonder just how big Big Data is? Check out this Mashable article on the lengths a woman went to in order to hide knowledge of her pregnancy from the Internet. It went a lot, lot further than just not mentioning it on Facebook. Good morning, Internet…

I love your comments, I love your site suggestions, and I love you. Feel free to comment on the blog, or @ResearchBuzz on Twitter. Thanks!

Guggenheim Exhibition Catalogues, Now Digitized

Hat tip to The Spectator for the article about a digitized collection of exhibition catalogs from the Guggenheim Museum. The collection is available at

There are over 60 catalogs here, and when you sort by date you’ll start with “Amazons of the Avant-Garde” (1999) and end up at “Art of Tomorrow: Fourty-One Reproductions from the Solomon R. Guggenheim Foundation for Non-Objective Painting” (1940). The listings contain a thumbnail of the catalogue cover, author, date of publication, and number of pages.

I clicked on the “More” listing for “Mastercraftsmen of Ancient Peru,” by Alan R. Sawyer (1968, 112 pages). I got a larger image of the cover and a brief excerpt of the book, with a “Read Catalogue Online” link. THAT took me to a Flash-based reading application. Navigation of the book at the bottom, double-click to zoom in (to read the text you will have to zoom in.) Double-click again to zoom out. When I was looking at it, it looked like the reader was on “auto-play” — so you’d be looking at some page and it would flip. You can turn that off by clicking the “pause” button on the nav bar. If you don’t want to bother with the nav bar you can also flip through the pages by clicking on page corners.

The bottom of the listing has books related to the catalogue you’re viewing as well as related essays. I’m not sure where “Aestheticism and Japan: The Cult of the Orient” intersects with ancient Peruvian crafts, but I can find out if I pay $1.99 for this 13-page ebook.

An absolute timesink. If you don’t want to do the reading on the Guggenheim site, you can download a large selection of texts — more, it seems to me, than there are at the Guggenheim site — at the Internet Archive. This includes downloading in Kindle, Daisy, and PDF format.

Stained Glass Windows of Wales

Robert Jones-Morris
Robert Jones-Morris

This one has been sitting in the queue for a while (I have continued checking my information traps even when I haven’t had time to write about anything) and it’s gorgeous! A story in WalesOnline alerted me about a resource cataloging stained glass windows in Wales … over 5000 photographs of about 1800 windows from 350 locations across the country.

The archive is available at You can browse by artist or location, or search by keyword. Being stained glass windows, the archive is pretty much all religous-themed; a search for angel finds 823 results. A search for grail, however, finds four. An advanced search lets you find windows by the date they were created. I found some windows going back to the 15th century.

Individual entries vary a lot. Usually there’s a picture, a link to see a larger version, and a brief history of the window and its location. On the other hand, the entry for “Christ in Majesty with St Michael and Angels” had several details about its history as well as many, many detail photographs including a detail of Robert Jones-Morris, who makes an appearance at the bottom of the window, playing the organ.

A beautiful collection of art. If you like what’s here, check out the Harry Ransom collection of medieval and early modern manuscripts, which have amazing illustrations.

FBI’s Online Database of Stolen Art

With the recent redesign of, the Federal Bureau of Investigation has launched the National Stolen Art File, as it noted last week. This new database contains information about thousands of stolen art items across the US, from books to paintings to stamps, even. It’s available at

(Before you get into searching, read the site’s description of what is qualified for listing in the file: “The object must be uniquely identifiable and have historical or artistic significance.” That’s a pretty wide mandate. Also it has to be worth at least $2000 “or less if associated with a major crime.”)

The search form on the right allows you to specify the type of object — from Altar to Wine Cooler — with a few other fields including title, maker, period, and “additional data,” which I think is a full-text keyword search. I specified just that I was looking for information on stolen books, and didn’t fill out any other fields.

I got around 300 results in groups of 20. Result pages include the name of the book and the copyright (or period date; I don’t think anything’s copyright 1366.) Sometimes a thumbnail image of the item is available. (The FBI announcement of the NSAF launch notes that there are about 7,000 images in the database right now.) Clicking the item gives you a spaces for some additional information, like dimension and materials, but descriptive information varies a lot.

The books were sometimes but not always particularly old; The Method of Manjusri’s Secret Yoga for Observing the Self and Becoming a Buddha in One Chapter, produced in China in 1316, was next to Speech of Acceptance Upon the Award of the Nobel Prize for Literature (William Faulkner, 1951) and near Walt Whitman: a selection of the manuscripts, books and association items gathered by Charles E. Feinberg (1955.)

The most surprising omission in the database listing was any kind of date of listing or notation on when the item was added to the database. I can also imagine that an RSS feed of new items in specific categories would be useful — if I were an antique dealer, for example, it’d be good to have an RSS feed of antiques reported stolen to the FBI, so I could stay aware of stolen items.