hello I am a bit new to this,
is it possible to take advantage of python's machine learning libraries with in Poser?

does anyone have any idea?

I assume you want this for 2D graphics processing via AI-generated or manipulated images. I suspect the ideal solution would be to integrate the free GMic into Poser 12's PostFX, since GMic (aka G'Mic) now has good machine-learning inputs and abilities. If Poser 13 lets us plug any Photoshop filter into PostFX (GMic is now also a Photoshop filter), and send it Python commands via a script then that might be a start? Just guessing on this.

@ HartyBart: check your settings.  You show up as blocking me from your comments.

Hi Hartybart, thanks for your answer, I don't know exactly what Gmic is for, but there are some python and machine learning libraries that allow low level image editing.
I believe that in the not too distant future it will be possible to make a fast rendering of an image and increase its resolution by using AI.

GMic is a bundle of adjustable and powerful image-processing filters, available in many formats including as a free Photoshop plugin. It is open-source and has recently built in a powerful machine-learning library called the 'Neural Network Library' ('nn_lib' and related sub-modules and calls), which is being coded from scratch and is now about a couple of years old.

GMic 3.x public release currently includes a denoiser based on the machine learning library. However... if all you want is speed via AI denoising, then the AI-based Intel 'OIDN' CPU denoiser is already in Poser 12 and is excellent. It allows the rendering of a BIG 'quick and grainy' render, which it can then clean up without too much loss of detail. This greatly speeds render times on SuperFly (Cycles). It also has the advantage of outputting two renders, normal and denoised, which means you can load them as Photoshop layers and then erase or blend any bits that are too smoothed out.

I seem to recall that there's also a graphic-card based equivalent denoiser in Poser 12, but I have lots of cores and so I use the Intel CPU one.

Note also that Poser 13 - due fairly soon - will have the new Blender Cycles which is said to great speed up render times for animations.

Poser 11 lacks built-in denoising. But for P11 users on Windows, there are two freeware Windows GUI versions of Intel’s CPU 'OIDN' Denoiser, for desktop PCs. Obviously, you run the render through these separately after the Poser render is made. Although a custom PoserPython script would give you a one-click: "render with Poser, and then call another .EXE and use it to process and save the render". That said, recent Windows OS's may well freak out at the user trying to do that, re: security. But it can be done.

There's one thing to remember. Denoisers may not be 'temporarly stable'. In plain English this means that animators may see flickering or wobbling between animation frames, even when they have been processed with the same denoiser.

GMic comes with Krita. Krita also seamlessly reads the raw (optionally HDR) render files Poser produces. 

A great opportunity for AI in Poser would be poses. There is a huge library of poses available that could be used as learning material and it would be great to be able to tell a scene figure what to do rather than go in and bend it like a doll.

Another application for AI would be animations that do not let hands pass through body or legs. This is needed for dynamic clothing.

Python is a popular language for machine learning. You can use libraries like NumPy, Pandas, Scikit-learn, and Matplotlib to perform machine learning in Python. The basic steps are importing the dataset, preprocessing it, splitting it into training and testing sets, training a machine learning model, evaluating its performance, and making predictions on new data.

Click here more information: website development solutions