Timelapse Mining from Internet Photos

An der University of Washington haben sie eine Software gebastelt, die Fotosammlungen im Netz scannt und daraus automatisch Timelapse-Videos von Orten generiert. Macht das langweilig gewordene Zeitraffergedöhns auch nicht spannender, aber tolles Projekt:

YT: Time-lapse Mining from Internet Photos (via Quartz)

We introduce an approach for synthesizing time-lapse videos of popular landmarks from large community photo collections. The approach is completely automated and leverages the vast quantity of photos available online. First, we cluster 86 million photos into landmarks and popular viewpoints. Then, we sort the photos by date and warp each photo onto a common viewpoint. Finally, we stabilize the appearance of the sequence to compensate for lighting effects and minimize flicker. Our resulting time-lapses show diverse changes in the world’s most popular sites, like glaciers shrinking, skyscrapers being constructed, and waterfalls changing course.

Anagramatron: Tweets zu Anagramme


Anagramatron ist ein kleines Script, das auf Twitter nach Anagrammen sucht, sprich zwei passende Tweets zusammenstellt. Zum Teil kommen da fast schon lyrische Paarungen raus, nachzulesen im zugehörigen Tumblelog oder direkt bei @anagramatron(via coudal)

This script connects to the twitter stream. When it receives a new tweet it runs it through some filters, ignoring tweets that contain things like links or @mentions, or that contain less then a minimum number of characters. It then sorts the characters in the text in alphabetical order, ignoring non-alphabet characters; this ordering serves as an anagram-unique hash, i.e. any pair of anagrams sorted this way will produce identical strings. This hash is checked against a list of all the hashes we have stored so far. If nothing is found, the hash and the original text are saved in a database. If a match is found, the original text of both tweets are run through some comparison tests to check for like-ness. If they pass that text they are flagged for review, to make sure they aren’t too similar, or haven’t been posted previously, etcetera…