Template talk:Feature detection (computer vision) navbox
This infobox steals valueable article space, at the top right of the screen, which could be much better used by illustrations of the article's topic. Please replace this infobox with a navbar at the bottom of the article, after the "See Also" section. See Template:Tuberculosis for example. (Consider that readers who get to an article on X will most likely want to read about X, not about other topics that someone thought are connected to X.) --Jorge Stolfi (talk) 21:46, 22 August 2010 (UTC)
I understand this concern. For some of the feature detection articles, this navigation box is really useful and has been included by a number of different authors of feature detection articles. For other feature detection articles, like the structure tensor that I assume that you refer to, there might be other priorities? A solution that would satisfy both audiences could therefore be to develop a corresponding navigation box to be put at the bottom for such articles I'll try to find time to make such a navigation box. For previous feature detection articles, however, it seems preferable to keep the current layout to simplify navigation between different feature detection topics. Tpl (talk) 14:13, 23 August 2010 (UTC)
Now this navigation box is much wider than it used to be. Does that have to do with the most recent change on February 13, or could there be some other reason, e.g. changes in the underlying templates? Tpl (talk) 14:34, 11 March 2012 (UTC)
Undesirable promo of COSFIRE
The intention behind this template is to list the most well-known and commonly used feature detectors in computer vision. The recently added COSFIRE operator under the header interest point detector is, however, not an established interest point detector in the field. To me, the addition of this operator therefore appears as an undesirable self promo.
I would therefore suggest removal of this interest point detector from the template. Previously, I removed it with a comment, but it was immediately put back by the person adding it. To avoid repeated edit conflicts, I would therefore suggest a wider discussion about this. Tpl (talk) 08:34, 9 May 2012 (UTC)
Should LESH be removed from list of feature descriptors?
All the other algorithms in this list are now considered classics on computer vision and widely cited and discussed in books/papers. LESH is relatively unknown. A search on google scholar only shows 35 citations for "Head Pose Estimation in Face Recognition across Pose Scenarios", while the SIFT and SURF papers about 17,000 and 2,600 respectively. If every descriptor that has been proposed in the past 10 years is added to this list it will be huge. Pabeles (talk) 23:44, 30 November 2012 (UTC)