-
-
Notifications
You must be signed in to change notification settings - Fork 2.2k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We鈥檒l occasionally send you account related emails.
Already on GitHub? Sign in to your account
clarify the meaning of the weights term for TV denoising functions #6134
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thanks for your contribution, @gmelon.
As is, the only addition brought by this changeset is saying that 'more denoising' means 'more smoothing' -- it's not addressing #5379 (and since a noisy signal is notably not smooth, the reader already understands intuitively that 'more denoising' means 'more smoothing').
In the original issue, @grlee77 pointed out that what needs clarifying is which term in the cost function the weight applies to. And I would argue that the double negative form "the expense of less similarity to the input
" calls for rewording.
@mkcor, Thank you for the comments. In summary, do you mean that "the expense of less simplicity to the input" needs to be modified in some way? |
Ideally, we would add the actual cost functions using LaTeX math markup similarly to the example here: scikit-image/skimage/metrics/simple_metrics.py Lines 196 to 202 in d52c11d
The user can then see the equation that is being minimized and where the Both functions are solving a similar problem, but the equation is formulated a bit differently in the two publications. If I recall correctly, in one of them the weight enforces similarity to the existing image whereas in the other the the weight is applied to the TV-regularization term in the equation. So, a larger |
@gmelon no, I said three things, so they cannot be summarized into one. I said:
Please let us know if you run into any roadblocks! |
Hi @gmelon, |
Thanks for asking! I'm sorry to say that I can't continue this issue now. Sorry again and Thanks you! |
Description
Fixed #5379
After discovering above issue, I modified the docstring to clarify the meaning of weight a little more.
Please review it. Thanks you馃檪
Checklist
./doc/examples
(new features only)./benchmarks
, if your changes aren't covered by anexisting benchmark
For reviewers
later.
__init__.py
.doc/release/release_dev.rst
.