Jump to content

AI tools coming to Affinity products?


Recommended Posts

Hello,
Is there any AI features planned for Affinity products?
I would like to see AI masking features like we can see in Capture ONE. This would save a ton of time when doing edits.
When doing for example portrait edits it would be really beneficial to be able to quickly select for example skin of the model for editing.
Detailed masking using brushes is a lot of time wasted. Hoping to get AI tools to help.

Regards,
Mika

Link to comment
Share on other sites

  • Staff
30 minutes ago, leinmi said:

Is there any AI features planned for Affinity products?
I would like to see AI masking features like we can see in Capture ONE.

Machine learning tools like those for making selections are being investigated, yes. We are not looking at add Generative AI features, but are looking at ways to improve workflows

Patrick Connor
Serif Europe Ltd

"There is nothing noble in being superior to your fellow man. True nobility lies in being superior to your previous self."  W. L. Sheldon

 

Link to comment
Share on other sites

A big thumbs up to AI/ML-based selections. Automating two basic types of selection - select sky and select object - would make masking and hence workflow so much smoother. Imagine a photo like the one below, and how difficult it is now to select just the sky, or just the tree. Streamlining these two basic selection types would make me so very, very happy.

IMG_1370.jpeg.795006444b1017ea2daf80a30149076f.jpeg

Affinity Photo 2, Affinity Publisher 2, Affinity Designer 2 (latest retail versions) - desktop & iPad
Culling - FastRawViewer; Raw Developer - Capture One Pro; Asset Management - Photo Supreme
Mac Studio with M2 Max (2023); 64 GB RAM; macOS 13 (Ventura); Mac Studio Display - iPad Air 4th Gen; iPadOS 18

Link to comment
Share on other sites

.....and masking out HAIR and other fine, wispy details. 👍

2017 15" MacBook Pro, 16 GB RAM, Ventura v13.7, Affinity Photo/Designer/Publisher v1 & v2, Adobe CS6 Extended, LightRoom v6, Blender, InkScape, Dell 30" Monitor, Canon PRO-100 Printer, i1 Spectrophotometer, i1Publish

Link to comment
Share on other sites

16 hours ago, Patrick Connor said:

Machine learning tools like those for making selections are being investigated, yes. We are not looking at add Generative AI features, but are looking at ways to improve workflows

Thank you, there are many, many hours to be saved by actually working faster and with higher quality in the results. Can you answer whether AI is being looked at for Designer, specifically for vector features? You don't need to go into detail, but there are also many time-consuming workflows that could be reduced with AI. And I am NOT thinking of a vectorizer, but all the usual workflows and the quality of the output.

Festina lente

Link to comment
Share on other sites

  • Staff
29 minutes ago, Aldus said:

whether AI is being looked at for Designer, specifically for vector features?

Initially we're focusing on Photo, but I'm allowed what our future plans are, sorry.

Patrick Connor
Serif Europe Ltd

"There is nothing noble in being superior to your fellow man. True nobility lies in being superior to your previous self."  W. L. Sheldon

 

Link to comment
Share on other sites

AI selection would be excellent indeed. I hope the devs will take a look at AI upscaling as well. Would be pretty nice for restoring low res images to be of higher quality.

As for other AI features like fill generation, that stuff could be fixed once the plug-in API gets finished. A strong API will open the floodgates for a lot of stuff, including AI features.

Link to comment
Share on other sites

21 hours ago, Patrick Connor said:

Machine learning tools like those for making selections are being investigated, yes. We are not looking at add Generative AI features, but are looking at ways to improve workflows

This is good news. We don't need generative AI tools to Affinity products, but AI/ML-based masking and selections would be great help for us. Just like Capture ONE has and ACDSee has AI masking tools too. Should be possible for Affinity...hopefully soon.

-Mika-

Link to comment
Share on other sites

23 hours ago, Patrick Connor said:

Initially we're focusing on Photo, but I'm allowed what our future plans are, sorry.

I think there would be some major potential for Publisher as well when it comes to copyfitting and tweaking composition. Such a feature could be used to help with making a text fill a frame for example, or to fill gaps between words that are visually too big.

Say you need another line of text in your frame and you click a command that offers a few good alternatives to choose from for adding filler words, very minor rephrasing etc. to the story. The proposed text changes could be shown in some sort of diff view or in-place in the layout in a highlight color with the option to accept each suggestion into the story or reject it. Conversely, if a story is a few lines too long, it could offer suggestions for which parts to cut first.

The part with the manual review of the suggestions is important because I don't think such a feature should just rewrite a text on its own without control, just speed up the existing manual workflow by offering options.

Link to comment
Share on other sites

52 minutes ago, Peter Werner said:
On 5/21/2024 at 2:35 PM, Patrick Connor said:

Initially we're focusing on Photo, but I'm allowed what our future plans are, sorry.

I don't think such a feature should just rewrite a text on its own without control,

Considering the difference that one word can make to the meaning of something, it's probably a very good idea not to let AI rewrite what you have written! 😉 (People can make enough mistakes on their own, without allowing AI to make more!)

Acer XC-895 : Core i5-10400 Hexa-core 2.90 GHz :  32GB RAM : Intel UHD Graphics 630 : Windows 10 Home
Affinity Publisher 2 : Affinity Photo 2 : Affinity Designer 2 : (latest release versions) on desktop and iPad

"Beware of false knowledge, it is more dangerous than ignorance." (GBS)

Link to comment
Share on other sites

9 minutes ago, PaulEC said:

Considering the difference that one word can make to the meaning of something, it's probably a very good idea not to let AI rewrite what you have written! 😉 (People can make enough mistakes on their own, without allowing AI to make more!)

It is a helpful tool already for composing emails and cleaning up writing. I am not for it any artistic sense, but for business I have no issue running an important email through AI and then go over it after. It is always on the user though to check the work of AI. I carefully read anything I have AI generate for me before sending it out. I have yet to have it misunderstand what I am trying to say and say something completely different. 

Link to comment
Share on other sites

27 minutes ago, wonderings said:

and then go over it after.

That's my point! Even if AI gets it right most of the time, I would still be very wary of using any AI generated text without carefully checking it! Unfortunately a lot of people would simply assume that the computer is always right; in the same way that some people follow a satnav blindly and end up in a canal! 🙂

Acer XC-895 : Core i5-10400 Hexa-core 2.90 GHz :  32GB RAM : Intel UHD Graphics 630 : Windows 10 Home
Affinity Publisher 2 : Affinity Photo 2 : Affinity Designer 2 : (latest release versions) on desktop and iPad

"Beware of false knowledge, it is more dangerous than ignorance." (GBS)

Link to comment
Share on other sites

9 hours ago, debraspicher said:

Is this a moral call? Just curious.

You can go to Twitter (X), and search for generative @affinitybyserif and you should find a relevant post (it's second on the screen when I tried it just now).

-- Walt
Designer, Photo, and Publisher V1 and V2 at latest retail and beta releases
PC:
    Desktop:  Windows 11 Pro 23H2, 64GB memory, AMD Ryzen 9 5900 12-Core @ 3.00 GHz, NVIDIA GeForce RTX 3090 

    Laptop:  Windows 11 Pro 23H2, 32GB memory, Intel Core i7-10750H @ 2.60GHz, Intel UHD Graphics Comet Lake GT2 and NVIDIA GeForce RTX 3070 Laptop GPU.
    Laptop 2: Windows 11 Pro 24H2,  16GB memory, Snapdragon(R) X Elite - X1E80100 - Qualcomm(R) Oryon(TM) 12 Core CPU 4.01 GHz, Qualcomm(R) Adreno(TM) X1-85 GPU
iPad:  iPad Pro M1, 12.9": iPadOS 18.1, Apple Pencil 2, Magic Keyboard 
Mac:  2023 M2 MacBook Air 15", 16GB memory, macOS Sequoia 15.0.1

Link to comment
Share on other sites

I would like selection AI too but not the one in Photoshop.   Honestly  I can't select a thing with Ai selection in Photoshop except cats, cars and a few other perfectly select-able without any AI things. 

It should be something driven by  description,   recognizing what I actually want to select  . For example a crack in asphalt  or a specific tree  in a forest . Otherwise it would be same annoying thing like in Photoshop  you rather waste your time on .  In cant even select a sky  with a clean horizon line .

Would love also de-lighting  AI , something removing shadows  from where you don't need them .   Best if not generating  new content within shadows  but rather  peroperly de-light the whole image  like some old  now defunct soft   https://www.cgchannel.com/2019/06/tandent-vision-science-launches-lightbrush-cloud/

Too bad I have it only on my old mac mini  and it's impossible to put on a new pc now.    Works great there  but takes forever on big images .  The whole company looks like dead . Not sure what happened to those guys invented it. 

 

As of generative AI in Photoshop  I still couldn't imagine what I might it need for . It always did something  blurry and instantly recognizable as a fake.

 

 

Link to comment
Share on other sites

On 5/22/2024 at 10:35 AM, PaulEC said:

That's my point! Even if AI gets it right most of the time, I would still be very wary of using any AI generated text without carefully checking it! Unfortunately a lot of people would simply assume that the computer is always right; in the same way that some people follow a satnav blindly and end up in a canal! 🙂

I guess that is the different between using AI to enhance your work and using AI to do ALL the work. 

I am looking at getting a Tesla Y and really want the self driving option as my hour long commute would be a bit more enjoyable with it. I do understand though that it is not something I turn on and then fall asleep at the wheel only to wake up when I get to work... or end up in a ditch or worse! Tools to help make things easier yes! Tools to do all the work for me, no!

Link to comment
Share on other sites

I regard using AI as like using Google translate.

For example, please consider this song in Slovenian.

I can find the lyrics on the web.

I can use Google translate to translate the Slovenian lyrics to English to get an idea of what the song is about.

Because i can use my knowledge of English, my native language, to consider what Google translate presents to me.

Yet I would not use Google translate to publish a Slovenian version of one of my song lyrics.

Likewise when I asked Bing Chat ai to produce a painting of a lady in a long green dress feeding an okapi I could assess each of the pictures produced for me as to whether the animal depicted was like an okapi. In some of the pictures, the likeness was very good, in others it was way off. So I chose to go forward with what I considered a beautiful painting and buy a print and frame it, yet not using the other pictures. So, in my opinion, generative AI an be very good, provided one checks carefully the results that are provided.

William

 

 

 

Until December 2022, using a Lenovo laptop running Windows 10 in England. From January 2023, using an HP laptop running Windows 11 in England.

Link to comment
Share on other sites

A problem with the policy of not allowing any AI generated content to be posted in the Affinity forum nor linked to from a post in the Affinity forum is that I cannot display a painting generated using AI and ask how I can achieve such an effect starting with a blank canvas in Affinity Designer using only the tools in Affinity products.

William

 

Until December 2022, using a Lenovo laptop running Windows 10 in England. From January 2023, using an HP laptop running Windows 11 in England.

Link to comment
Share on other sites

On 5/24/2024 at 6:23 AM, William Overington said:

I cannot display a painting generated using AI and ask how I can achieve such an effect starting with a blank canvas in Affinity Designer using only the tools in Affinity products.

Why do you need to have a painting generated by AI as your starting point?

If you've asked the AI to create a painting in some style, just find an actual painting online by a human artist using that style, and use that as your model to learn from. There are plenty of Stock sites with images you could ask about, or museums with collections online, etc.

-- Walt
Designer, Photo, and Publisher V1 and V2 at latest retail and beta releases
PC:
    Desktop:  Windows 11 Pro 23H2, 64GB memory, AMD Ryzen 9 5900 12-Core @ 3.00 GHz, NVIDIA GeForce RTX 3090 

    Laptop:  Windows 11 Pro 23H2, 32GB memory, Intel Core i7-10750H @ 2.60GHz, Intel UHD Graphics Comet Lake GT2 and NVIDIA GeForce RTX 3070 Laptop GPU.
    Laptop 2: Windows 11 Pro 24H2,  16GB memory, Snapdragon(R) X Elite - X1E80100 - Qualcomm(R) Oryon(TM) 12 Core CPU 4.01 GHz, Qualcomm(R) Adreno(TM) X1-85 GPU
iPad:  iPad Pro M1, 12.9": iPadOS 18.1, Apple Pencil 2, Magic Keyboard 
Mac:  2023 M2 MacBook Air 15", 16GB memory, macOS Sequoia 15.0.1

Link to comment
Share on other sites

1 hour ago, walt.farrell said:

Why do you need to have a painting generated by AI as your starting point?

It is not a need, it is a want. Not essential, just something that I want to do.

1 hour ago, walt.farrell said:

If you've asked the AI to create a painting in some style, just find an actual painting online by a human artist using that style, and use that as your model to learn from. There are plenty of Stock sites with images you could ask about, or museums with collections online, etc.

Because the AI outputs a 1024 pixel by 1024 pixel images, and some of them, in my opinion, albeit not a professional art expert opinion, beautiful pictures. So as I can observe that the AI can produce such 1024 pixel by 1024 pixel images as electronic images, I wonder if I can do the same. For me, that is not the same as the alternative that you suggest.

I have tried to make a start.

Here are two threads.

and

 

William

 

Until December 2022, using a Lenovo laptop running Windows 10 in England. From January 2023, using an HP laptop running Windows 11 in England.

Link to comment
Share on other sites

  • 3 months later...

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
×
×
  • Create New...

Important Information

Terms of Use | Privacy Policy | Guidelines | We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue.