You are currently browsing the Thoughts of a Bohemian weblog archives for the day April 2, 2008.
- alexa (5)
- Aurora (2)
- Canada (8)
- celebrity (34)
- CEPIC (17)
- Cnn (3)
- commercial stock (32)
- copyright (31)
- corbis (92)
- Cosmos (1)
- digg (4)
- editorial (153)
- filter (17)
- finance (45)
- flickr (54)
- focus (13)
- france (22)
- getty (139)
- google (24)
- gumgum (9)
- HOLGA (8)
- idee (8)
- IPTC (13)
- Jupiter (21)
- keyword (39)
- law (24)
- lens (14)
- lensbabies (8)
- license (63)
- magazine (62)
- Magnum (6)
- mediastorm (13)
- Microstock (91)
- Midstock (22)
- msnbc.com (10)
- multimedia (36)
- news (75)
- newspaper (35)
- Newsweek (4)
- No sense (29)
- PACA (14)
- Pacific coast news (3)
- photojournalism (96)
- Photoplus (2)
- photoshop (6)
- Piclens (2)
- pictogram (1)
- picturemaxx (1)
- Plus (3)
- prosumer (40)
- Royalty free (64)
- Search (51)
- SIPA (6)
- slideshow (29)
- technology (78)
- TIME (13)
- transaction (55)
- Uncategorized (20)
- web 2.0 (75)
- wire service (20)
- yahoo (10)
- Zymmetrical (6)
- November 18, 2008: An Open Letter to Mark Getty
- November 9, 2008: A piece of fettuccine making it's way to an Alfredo sauce
- November 5, 2008: Photography and Petanque
- October 31, 2008: Dirty laundry
- October 26, 2008: "This is our company together.”
- October 23, 2008: The princess's price
- October 18, 2008: Picture this: Berliner and Rex merge to take on US image market
- October 11, 2008: The end of the stocker
- October 7, 2008: BollyPhoto
- October 3, 2008: one, two, three..any one else ?
Blogroll
Important Destinations
Subscribe Here :
Archive for April 2, 2008
photography and farming
April 2, 2008 by pmelcher.
A new company emerged from unknown depth a few days ago, proposing free “automated” tagging, or keywording. Named Tagcow, the company does not explain how the tagging is done.
Curious, I decided to give it a spin with a couple of images including this image:
After two days, the image was finally tagged with two words: “Pool” , “Man”. I guess that is what you get for free. Furthermore, when I downloaded the image, I could not find the keywords anymore…There is nothing automated about this service. It is currently impossible for a computer to recognize the content of an image. The most advanced systems I have seen have a 10% success rate, and then again with very contrasted and simple pictures. Tagcow uses a little known service offered by Amazon call the Mechanical Turk. With this service, anyone can put a long, painful task and offer to pay for human beings to fulfill them
“Complete simple tasks that people do better than computers. And, get paid for it”. At 0.01 cent a picture, one can get images keyworded for cheap. However, the quality is not guaranteed.
Thus it is the power of the masses used here, making more obvious why they picked a cow for their name. Not the brightest animal in the land
Posted in No sense, Search, copyright, IPTC, keyword, flickr, prosumer, web 2.0, filter | Print | 2 Comments »


