itch.io

In current version, the image going to place on the center of detection box. This gave me an idea to adjust image position with (X,Y) position. Such as "quadrant" with maximum for each quadrant is 1. 

Therefore, I can place the image in (-0.5, -0.5) from the center which is a lower left of detected box.

I think this give a possible to place image like QOS picture. (Also, enable to adjust picture rotation is gonna be good too.

The AI version will be amazing, but it’ll take a lot of time because of model training and prompting. You could release it as a DLC, pretty sure no one would complain.

Thank! However, I often found false negative on Asian male. Could adding model to pre-detect a male without using only a face? 

Anyway, Thank you for the great work!

As version 0.7 containing a bug about image is missing for brothel when enter a room.

Also, great work! Keep going bro.

Yep. I mean exposed skin. 

It possible to censor only visible part of human?

Due to a size of pixelate is constant which mean if a censored part is have higher area -> less pixelate.

It possible to make a pixelate size to be coefficient then calculate a detection area multiply by a coefficient value.

Thank you PerfextFox265! This product is awesome.

I find a bug when open a program, pixel censor type with low censored scale on a face part is not apply as correct scale. I have to manually rescale again. This occurred every time I open a program.

Sometime, I want to chillout and not want to sit in front of PC/Laptop. If you build the version for MacOS it will be great!