I’m just asking about the “regular” text – not the large, ornate stuff.
What would be the best way to go about this?
(I also was going to post what the results of AI Mask’s “subject” finder look like, but cannot do so. But the results are like it’s trying to find one big subject in part of the image.)
Place Green (Keep) on the large composites and then the strokes you see here in Green on the black, only the ones you see on this screenshot … i.e. 6 strokes on the black text in this cluster
Using a Mask Mode of AI compute the mask, then you will need to refine using the Color Range tool.
As for object detection using Subject, there are none in your image.
Hi, @AiDon. Thanks so much. I was enthused to wake up to a response so quickly.
I went through the steps you laid out. After punching the button to compute the mask, some things happened (I infer) – there was a message about GPU (“one-time performance benchmarking”), then the same one about CPU, then just benchmarking… I didn’t just sit and watch it the whole time. But when I came back to it minutes later, everything within the outer application window was white. I waited a few more minutes, while also searching this forum. But it was still like that. I supposed by that point that this was not intended white – that it crashed or froze.
I’ll check the preferences I saw you reference to others, and make sure it’s just on CPU. (I don’t have an external GPU – only integrated graphics.) But I’m reporting the problem I came up against all the same (and will report back about my second shot at this).
In Preferences, I selected CPU and I hit calibrate. I went through the steps again that @AiDon laid out above. Then I saw the same GPU and CPU-related messages. Then it froze again – on 11%, and not with all-white inside the app window.
I quit and will try again. This time I will not hit calibrate and see if that makes the difference.
It went immediately to “Processing Mask: 11%”. (Actually, it did that the 2nd time, too – but then it went into the GPU and CPU benchmarking; then back to “Processing Mask: 11%”.) This time, it’s just frozen here at the supposed 11% (for 10-15 minutes now).
I will quit, then look briefly again in this forum for similar issues with solutions. If I find a promising one, I’ll try it and report back. Otherwise, I will just be waiting and seeing if someone can help me further here.
Okay. After looking at a different thread in the forum, I thought I had better double-check the system requirements. Here’s what I have:
OS: Windows 10 Home, 64-bit RAM: 8GB Monitor: Resolution = 3200 x 1800 x 59 hertz openVINO: don’t know what that is (could it possibly help?) CPU: Processor = Intel(R) Core™ M-5Y71 CPU @ 1.20GHz, 1401 Mhz, 2 Core(s), 4 Logical Processor(s)
At least I think that last one’s an . Could someone confirm?
Disappointed if that’s the case. And my apologies if these posts have wasted people’s time. I thought I had checked them, but perhaps those were for a different Topaz app.
Is there is any possible way for me to run Mask AI given the above? (If it’s slow but results are good, I’m fine with that. )
Thank you. Having selected CPU before my 2nd attempt, I thought it would still be selected subsequently. So I surmise in hindsight that also pressing the “calibrate” button at that time resulted in CPU no longer being the setting later.
I selected CPU and got some output. I’ll have to keep experimenting in pursuit of better results. But I got output I could see and save.
Meanwhile, it seems that this aspect of Mask AI – initial use and calibration – that led to multiple crashes could stand for improvement (at least for a system like mine). But most importantly, I’m very pleased for it to be functioning now.
They are even confusing to me sometimes. Eg, I will use the green keep marker, and the very pixels I have gone over in green aren’t kept. That doesn’t make sense to me.
Given all that it seems like I would have to do by hand to get a quality mask, I am wondering about a different approach.
I can easily make a rough mask (binary image) of my original image using python/OpenCV or a program I’ve been utilizing from github. The result is not perfect (or I wouldn’t be here, of course. ); but it is decent. I could take the result and apply morphological operations – open and close to get rid of small holes, and then erode, to make the objects a little smaller. If could take the result of this and somehow use it so that the white portions (which are objects) could be translated or assigned to green/keep within Topaz AI Mask, that would seem like it would go a very long way, providing huge amount of correct “keep” marking for AI Mask to go on – and hopefully translating into much better results (and faster).
Is this possible? Can one input a binary mask image (or something like it) in order to provide a great starting point for AI Mask to then do it’s work?
OS: Windows 10 Home, 64-bit
openVINO: don’t know what that is (could it possibly help?)
CPU: Processor = Intel(R) Core™ M-5Y71 CPU @ 1.20GHz, 1401 Mhz, 2 Core(s), 4 Logical Processor(s)
