Treadco’s first product. Intelligent Image Sharpening.

Recently we announced Treadco – bespoke machine learning.

It looks like software in general will be our product.

My … collaborator has a rather rough way of expressing himself. It made me blush, but he said it was time to pull something out of somewhere. As I said, he’s coarse and uncouth.

Anyway, our first open source teaser product will sharpen images. We’ve discovered the joys of numpy and python – Fourier transforms that use the GPU and handle normalization like a breeze. It means we can easily, in something like 30 lines of code, implement a regularized Jacobi solver to find the image and point spread function that provides an optimal sharpening for a degraded image.

The application we’re looking at is in cryo-EM, but my collaborator mutters about stupid biologists and confocal microscopy. (I’ll have to get him drunk and find out what he’s pissed about. It must be something – he’s usually straightforward about these things.)

For those of us who remember our numerical analysis (a rapidly vanishing group of computer scientists) Jacobi iterations are an easy way to solve large matrix problems. It’s not quite as efficient as Gauss-Siedel and SOR, but it can be implemented with the Fourier transform. Technically, regularization is a successive under-relaxation rather than over-relaxation.  Most importantly, it’s a block iteration and can be implemented with high efficiency on all sorts of computer hardware. Careful regularization of the iterations, using noise levels in the image (the noise estimates don’t have to be stationary) results in an excellent and stable sharpening of the image. Without all that dangerous messing about with unsharp masks and inverse filters.

Anyway, Treadco has opened a github repository (empty for the moment) and is starting to produce.

By the way, there is no connection, none, absolutely none, with any Georgia State University resources.

Birth of a new business.

Very preliminary, but we’re announcing Treadco – bespoke machine learning and datamining for the discriminating user.

The name and form might change, but we specialize in highly efficient and highly accurate machine learning and data analysis.

  • Generative models such at the Restricted Boltzmann Machine. Our algorithm is 14 or more times faster than the standard approach. This makes it fast enough to implement in javascript or other interpreted languages and still get results in near real time. Web-based data analysis anyone?
  • Fuzzy and probabilistic/possibilistic modeling. Correct handling of uncertainty in both the independent and dependent data improves the robustness of the models.
  • Big supervised or unsupervised data problems.
  • Model development. Not sure what’s best? Ask us.

Untangling the mess.

My close collaborator and writing partner and I were thoroughly excited to be part of Booktrope. We weren’t quite in the “last fatal wave,” but were darn close. There were clues, plenty of clues, that something was rotten in Washington (the state, not DC. Something’s always rotten in DC.)

For various personal reasons I let him take the lead. This meant letting my “Author Platform” lapse (at the company’s advisement). A very bad idea, and one that we’re both trying to rectify.

What’s worse is our work is in limbo because of creative teams. Neither of us want to stiff them. We had three books in the hopper.

One, “The Curious Profession of Dr. Craven” made it out, with a thoroughly botched launch. Almost to the minute, the moment we got our ducks in a row with an enthusiastic marketing person and a plan, Booktrope pulled the plug.

Another book, “A designing woman,” is editing limbo. I need to figure out if we can afford to continue paying the editor. She’s very good.

The third book, “The art of deception” is free and clear. No one but us worked on it, so we have no one else to bother with. Other than short excerpts it’s not been published, so it’s been sent the rounds of open publishers. We’ll see.

Youth Protection.

I just finished redoing my biennial scout leader youth protection training. Well worth doing if you have children or work with children. It has changed since I last took it to reflect changes in the legal responsibilities and the threats.

  • No one on one contact and two-deep leadership. Remove the opportunity to be an abuser.
  • Respect privacy. Changes include confiscating a camera or electronic device that might have been used to take inappropriate pictures, as well as reporting it to the scout executive asap.
  • A big section on cyber-bullying and similar things.
  • An expanded section on ‘normal’ (as if there was such a thing) bullying. 

All in all improved from two years ago and focused on how to keep both leaders and scouts safe.

http://myscouting.org will get you there.

What counts as a publication?

I’m in a bit of a quandary. I like to prerelease chapters and snippets. This is both to boost the eventual sales of the book and get feedback (not so much from this blog, but there are some writer’s sites that have been exceedingly helpful).

I’d also like to submit things for publication. So what counts? At least one Sci-fi magazine says anything, anywhere, anytime is pre-publication and therefore cause for rejection. It’s not so bad for self-publishing, but I’d hate to “Sc**w the Pooch” to borrow a military aphorism for other modes.

On the other hand having snippets lodged in the bowels of Google with a definite time-stamp that I can’t possibly alter does protect me from charges of plagiarism.