Talk:Bag-of-words model in computer vision

Context and clarity

This article seems to need some more context. Is this a widely-pursued avenue of research in computer vision? Or is it the product of a single research group?

• It is very widely pursued. It is the most common type of descriptor for entire images. Bag-of-Words models are sometimes used for more local representations, too (e.g. regions, bounding boxes) Jhhays (talk) 08:45, 10 January 2014 (UTC)

Many parts of the article are unclear. Just to pick a couple:

• In the introduction, it says "Here an image may refer to a particular object, such as an image of a car". Now, I'd think that "object" refers to the car, and that the image of the car is in fact the image. Which is exactly what you'd expect an image to be. Why the confusing statement about what "image" refers to?
• The "dictionary" in the representation section seems to be, in fact, an ordered list. The "dictionary" property of it (being able to set and access arbitrary keys, not just a sequence of consecutive integers) is not used at all. It's also unclear what this has to do with images, and it sounds like it's simply describing a particular clunky implementation of the NLP bag of words model instead of saying anything informative about it. (A less clunky dictionary would represent a histogram by mapping words to counts, as in {"John": 2, "likes": 3, "to": 2, ...}
  • Also, the counts as they appear in that example are incorrect.
• The article refers to several images that were removed because they had no license or source information. The author of these sections either needs to release the appropriate images under a free license, or rewrite the article to not depend on images.

Finally, the article needs to be more up-front about the fact that the features it is detecting are not, in fact, "words", especially given how it starts by giving examples that involve words. rspeer / ɹəədsɹ 21:52, 9 August 2008 (UTC)

It's definitely a widespread computer vision technique. I'm not sure I'd call it an active field of research, however, since it's pretty well understood (and surpassed) at this point.

• It's definitely an active research area. There is a steady stream of new work on feature quantization, compact features, how to include spatial information, etc. Jhhays (talk) 08:45, 10 January 2014 (UTC)

I'm hoping to fix up this article, once I have a bit of spare time. Ha ha, right. Mdockrey (talk) 00:46, 26 November 2008 (UTC)

This is still an active field of study, as in people still publish on this topic. Its not particularly nascent and as a matter of opinion not particularly fruitful lately. I edited for grammar because "Also the BoW model for object segmentation and localization is also lack of study," is not a sentence. Zoratao (talk) 18:54, 7 May 2013 (UTC)

I'm deleting the word "sparse" from the introduction. Bag of Words models are sometimes sparse. For instance, when doing large scale retrieval with an inverted index the Bag-of-words need to be sparse to make the index compact. But for many recognition tasks the histograms of visual words are not sparse at all. When using certain more advanced feature encoding schemes the histograms are very dense. Jhhays (talk) 08:45, 10 January 2014 (UTC)

This page fails to cite the single most relevant paper on the topic: Video Google from Sivic and Zisserman, 2003. http://www.robots.ox.ac.uk/~vgg/research/vgoogle/ The paper has nearly 3000 citations. Jhhays (talk) 08:48, 10 January 2014 (UTC)

Missing figure(s)

One section refers to a figure, but no figure is displayed within the document. — Preceding unsigned comment added by 67.186.198.46 (talk) 02:26, 9 September 2011 (UTC)

External links modified

Hello fellow Wikipedians,

I have just modified one external link on Bag-of-words model in computer vision. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:

• Added archive https://web.archive.org/web/20101203074412/http://www.vision.caltech.edu/malaa/software/research/image-search/ to http://www.vision.caltech.edu/malaa/software/research/image-search/

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

This message was posted before February 2018. After February 2018, "External links modified" talk page sections are no longer generated or monitored by InternetArchiveBot. No special action is required regarding these talk page notices, other than regular verification using the archive tool instructions below. Editors have permission to delete these "External links modified" talk page sections if they want to de-clutter talk pages, but see the RfC before doing mass systematic removals. This message is updated dynamically through the template {{source check}} (last update: 18 January 2022).

• If you have discovered URLs which were erroneously considered dead by the bot, you can report them with this tool.
• If you found an error with any archives or the URLs themselves, you can fix them with this tool.

Cheers.—InternetArchiveBot (Report bug) 17:47, 13 July 2017 (UTC)