Syncbot SyncPlayer software that recognizes porn movements

This is the place for all suggestions, releases and feedback regarding Milovana Hardware efforts.
Post Reply
Maxim7
Curious Newbie
Curious Newbie
Posts: 4
Joined: Sat Mar 18, 2023 3:42 am

Syncbot SyncPlayer software that recognizes porn movements

Post by Maxim7 »

>Syncbot is our most advanced AI masturbator with own developed neural network

Thing is that they have offline software that works on GPU or CPU and can accurately extract all movements from porn videos.

May be good idea can be to adapt their existing software and NN to make scripts for other devices? It'll save ton of time and allow to quickly make lof of such videos.
diglet
Explorer At Heart
Explorer At Heart
Posts: 203
Joined: Sun Dec 11, 2022 5:43 pm

Re: Syncbot SyncPlayer software that recognizes porn movements

Post by diglet »

https://syncbot.com/

It looks like their stroker has a rather small range of motion. The motors also aren't very large... It's an interesting design.

Last year I tried to automatically generate funscripts with motion detection using unsupervised algorithms (I wouldn't call it machine learning), I wrote an OFS plugin to do so. After just a few days of work it works well for basic situations, but the human scripter still has to clean up the data, scale it, select the major movement axis in the video and work on scene transitions. It ended up saving a bit of time but it needs a lot more work before it's as good as the other motion tracking plugin for OFS. My approach used optical flow and fourier transforms rather than object detection. This has some advantages (notably, it works even if the action is offscreen) but also some disadvantages.

When I looked into the problem, my conclusion was that it's really hard. They seem to have spend a lot of time and money on the problem, maybe my skepticism is unwarranted...
Maxim7
Curious Newbie
Curious Newbie
Posts: 4
Joined: Sat Mar 18, 2023 3:42 am

Re: Syncbot SyncPlayer software that recognizes porn movements

Post by Maxim7 »

diglet wrote: Mon Mar 27, 2023 2:57 pm When I looked into the problem, my conclusion was that it's really hard. They seem to have spend a lot of time and money on the problem, maybe my skepticism is unwarranted...
Yes, whole idea is not to try to make recognition, but instead to look at the data that it generates and write only simple converter to format used in scripts, so results can be used by people having other interactive devices.

Recently huge progress happened with NN, so I am sure their NN will progress fast also, if they will be financially reliable and sell lot of their products.
Post Reply

Who is online

Users browsing this forum: No registered users and 58 guests