How much money did the University of British Columbia actually earn from this patent?
I suspect that many patents are to stop others from patenting your work. It is a worry.
The patent on SIFT expired yesterday
11–20 of 69 posts
Re: The patent on SIFT expired yesterday
#12How much money did the University of British Columbia actually earn from this patent?
I suspect that many patents are to stop others from patenting your work. It is a worry.
Open publication (at least in western jurisdictions) serves as prior art. If you don't intend to patent, just write about what you're doing openly and often!
Re: The patent on SIFT expired yesterday
#13Is sift still relevant today?
> Is sift still relevant today?
It is extremely relevant.
The fact that there exist other methods that "consistently surpass SIFT in all reasonable metrics" is inconsequential. You can say the same thing about JPEG compression, but JPEG is not going to disappear anytime soon.
SIFT has very good performance (but there are better methods), it is reasonably easy to compute (but there are easier) and fast (but there are much faster). Its shortcomings are well-known and easy to understand. The combination of SIFT+Ransac is the bread and butter of image matching, and if you work on this problem you cannot seriously propose a new method unless you at least compare to it.
Re: The patent on SIFT expired yesterday
#14Earlier quoted context omitted.
I suspect that many patents are to stop others from patenting your work. It is a worry.
Open publication (at least in western jurisdictions) serves as prior art. If you don't intend to patent, just write about what you're doing openly and often!
Can any publication serve as prior art? Can I use a personal blog or do I need a certain type of publication?
I know of people who have sent copies of their work in the mail to themselves to prove prior art.
Re: The patent on SIFT expired yesterday
#15There have been many better feature detectors introduced in the mix over the past fifteen years. This change won't make too big a difference for people in the CV community.
Re: The patent on SIFT expired yesterday
#16Earlier quoted context omitted.
Open publication (at least in western jurisdictions) serves as prior art. If you don't intend to patent, just write about what you're doing openly and often!
Can any publication serve as prior art? Can I use a personal blog or do I need a certain type of publication? I know of people who have sent copies of their work in the mail to themselves to prove prior art.
Any publication is fine as long as you can provide evidence. In fact the only thing special about publication is that it is a convenient way to demonstrate evidence of the prior art. What could be tricky about e.g. a personal blog post is proving that it was published when you say it was and that time stamps are genuine.
Re: The patent on SIFT expired yesterday
#17There have been many better feature detectors introduced in the mix over the past fifteen years. This change won't make too big a difference for people in the CV community.
Mhhhhh, I'm a researcher in the field and I'd say it's a mostly inaccurate statement. The root-SIFT detectors and descriptors are still really good. If you try to match images in different conditions (day/night), then CNN based approaches are OK.
Which one were you thinking about?
Re: The patent on SIFT expired yesterday
#18SIFT works really well.
We use it for aligning input documents onto templates for OCR.
Super cool.
Re: The patent on SIFT expired yesterday
#19Is sift still relevant today?
There's a good paper on this: https://demuc.de/papers/schoenberger2017comparative.pdf "Our evaluation confirms that, as expected, learned descriptors often surpass SIFT on all evaluation metrics. However, we also observe that advanced versions of hand-crafted descriptors perform on par or better than the state-of-the-art learned feature descriptors, especially in the more complex SFM scenarios. As such, our paper dem…
This paper is a bit "old" by the way. An excellent paper was released a week ago : https://arxiv.org/abs/2003.01587
Findings are mostly the same. For day/day images, with a properly tuned pipeline, SIFT is really good.
Re: The patent on SIFT expired yesterday
#20More info about what this actually is: https://en.wikipedia.org/wiki/Scale-invariant_feature_transf... --- Also, reading the listing, I see 2020-03-06 - Anticipated expiration 2020-03-08 - Application status is Active Is "active" some kind of obscure reference with a nonintuitive meaning, or did the patent owner reapply for the patent?
You can't reapply for patents like that, but I don't know what that status means. Patents often get an adjustment to their term and last a bit more than their nominal 20 years (due to delays being issued, etc.). https://en.wikipedia.org/wiki/Term_of_patent