I used Gigapixel on the rocks because it did a slightly better job than HitPaw, but I used HitPaw for faces where it does the best job. Than in Photoshop I simply do simply masking to glue in the faces as it were and after that I applied a bit of blur and some film grain and contrast adjustment to make it all match a bit better. And as a final touch I did here and there a light touch up using heal and clone tools to remove the most distracting scanning debris. The image really needs to be properly retouched manually to be restored, but you can see how it could be done. I think., considering the starting position its pretty impressive what these tools can do. This is by the way just some crop of people in the picture, 4x enlarged from the original , here shown at 100% magnification.
If you are working professional and can see that as investment, it definitely saves you lot of time. For example dodge and burn manually for a face would take between 2-6 h for experienced retoucher. With this tool its 5 min. And results are virtually identical to a client.
A retoucher might not need all the plug ins. Also if you are working with studio images, cleaning up the floor automatically in batch saves a lot of time and with a few jobs you could make up for the entire collection much less one of the plug ins.
That being said, I agree it is expensive and its not for everyone or even every retoucher. For example there are some of the plug ins for removing red blood vessels in the eyes. I personally rarely have to do that, and its not a problem to do manually, so I would not buy that one. Someone who does a lot of beauty work and close ups or portrait headshots might need it more etc.
Thanks for your work on this! Looks very good considering the original, though my great-grandfather grew some glasses 
If you see the Mac version pop up, please alert me!
Hehehe. We can always add or remove glasses., hehehe. 
Sure, I’ll post probably here if I see Mac version. But overall I think its a company worth following, they are releasing about half a dozen products and I really like their approach and how they are trying to approach it. You can sense the company is more than just a cash grab operation. I have a feeling in the upcoming years they might be releasing some really good stuff.
Cheers!
If you are planning to use the face enhancement only, you can just use Tencent ARC’s GFPGAN. You can download the code from Github and it is free.
From what I have tested, HitPaw Face enhancement is identical to GFPGAN.
I am not convinced that those tools are actually offline.
The HitPaw photo enhancer actually says “upload” when you add a photo, so it might get sent somewhere.
Also, never forget, if you are using something for free, the service usually sucks up all your data including your images. I surely won’t be using anything Tencent touches, they are basically a part of the Chinese government.
Where does it say “upload”? I am pretty sure its not online, since I’ve done it the same way when internet is not activated. Actually their whole sales pitch is that they are not as much better than other tools , but they are offline and indeed as far as I can tell its offline. Unlike most online tools. That is what I don’t like about the other tools. Privacy concerns, slow and not very secure.
Tencent ARC’s GFPGA does look almost identical to HitPaw AI and what they are using. Makes me feel uneasy when I realize that since Tencent is essentially CCP (Chinese Communist Party) and they use mass surveillance on everyone, and that is how they are able to reconstruct faces. Once you realize that, its well, a lot like sausages… its best not to know how its made.
I always use window firewall to block all internet travel in/out of unknown software. HitPaw still working after I block it with firewall. 
Also I use “Microsoft Network Monitor” to check all network travel. I don’t see any unusual travel after I start HitPaw. 
Not sure where you see those word ?
I haven’t seen those word from my memory. 
The GPFGAN is open source. You can view those code in GitHub. 
Also it doesn’t require internet connection. Unlike Remini app or MyHeritage Photo Enhancer, which require to upload image to their server.
Take @plugsnpixels photo from above as an example.
Tencent ARC’s GFPGAN:
Tencent ARC’s GFPGAN:
Tencent ARC’s GFPGAN:
Excellent comparison. I think your instincts were correct. Tencent ARC’s GFPGAN is what HitPaw has added to their program. They just incorporated it into their interface and from what I can tell HitPaw’s Denoise Model is the same as face model, except in face model is essentially doing everything as denoise model, than does another pass with Tencent ARC’s GFPGAN and glues in the face that is better rendered. Its like bad Photoshop Job of composting since sometimes you can see the square box that was used to stitch the two models together.
Still, its a simple to use interface for us non coders, and it does the same job as I would sometimes have to do manually in Photoshop. And off course the Photoshop option is till available for best results, when multiple AI models are needed for best overall composite.
Cheers!
If the tech is open source and usable offline, I’m all for it.
Also, being open source would make it possible to integrate into Topaz software, wouldn’t it? Let’s see where this goes…
Great discussion!
So it seems the bottom line is, these apps replace features rather than enhance them. If the original is good, the results are good – maybe there is actual enhancement taking place for those. If not, we’re back to “Mr. Potato Head”, but pretty well done.
I would hope none of these results are ever attempted to be placed into evidence…
Thank you for sharing a new software for Photo Enhancment.
I think we better start a new Topic about this interesting software.
Great. I will reply there. Cheers!
That’s true. Anything requiring coding with no decent Graphical User Interface is out of the question for vast majority of users.
“On the PC, I still don’t know how good the app is because it refuses to generate a preview (“0%” flickers when I hit the Preview button but nothing happens). I also can’t see the two before-and-after panels you show (I see a single panel) or export the result to view separately when using the trial version. So that’s the end of that.”
Not sure about that. It always worked for me. Maybe you need to wait a bit for preview, but I never had an issue with it. I use it quote often, and I find it to be maybe best value for low light and low resolution smartphone shots of people. It cleans up the noise, enlarges 4x although its easy to down sample it to any resolution one needs, and best of all and most impressive it can reconstruction faces amazingly well even in difficult situations and the person is recognizable. I imagine this is the kind of tech they will start putting in the smartphones soon, because it really changes the smartphone photography.
PS. With the use of few other programs and plug ins in Photoshop its possible to regain some of the lost skin texture or generate new one, and also adding a bit of film grain, makes the faces not look as artificial with extreme cases.
By the way, there is a demo image that comes with the program, it demonstrates how blurry image can be reconstructed with their new Model C for faces.
ESRGAN is now much easier to use. Some newer GUIs (such as Cupscale) come with Python pre-packaged and allow for easy setup. There’s a new GUI in development that’s working on adding it as well. Topaz needs to step up their model game by quite a bit.
HitPaw, accepts only internet file types.
jpeg, png etc.
On my 17 Teraflop Radeon Pro W6800 it takes 3 seconds (2 seconds for resize) for 1% point.
So 5 minutes for one image to de-noise.
The face model did not change anything on an analog test photo.
You can’t maximize or enlarge the app window, it doesn’t accept drag and drop.
When denoising another image, there were tiles and leftover pixels that it didn’t recognize as noise.
Also all details were erased.
The app does not allow model changes with sliders, you can not set it to your own needs.
jpeg testimage: 2642 x 4177 and 5K.
Denoise Denoises 150 75Mb tiff images in the same time with this GPU.
Where it makes differences between noise and details.
So, how long will it take when the customes did not spend 1000$ to 4000$ on his GPU?
The photo apps of TopazLabs run quite snappy even on relatively small GPUs.
While the latest Gigapixel update is a big improvement over the previous versions and does many things really well, the face reconstruction is still lagging behind. Simple as that. You can complain all you want, but when you need that feature, Topaz simply cannot deliver… for now. Since all these are tools that require someone to get the job done, if the tool does not deliver, and other tool does, than you use the tool that delivers. I leave the fanboy stuff to others.
