EGAKU AIBêta
CreateEASYToolsGalerieTarifsJournalNouveautésDiscord
Accueil / Journal / Dev Log: Real Ukiyo-e, Ink Wash, Art Nouveau and Impressionism — Trained Only on Public-Domain Art
news6 min de lecture·2026-11-06

Dev Log: Real Ukiyo-e, Ink Wash, Art Nouveau and Impressionism — Trained Only on Public-Domain Art

How we replaced four anime style presets with style models we trained ourselves on public-domain prints and paintings, why stronger was not better, and two changes that cut extra arms and legs.

E
par EGAKU AI Team
dev-loganimeart-styleukiyo-eink-washart-nouveauimpressionismanatomy
Essayez-le sur EGAKU, gratuit pour commencer→

Why we did this

EGAKU's anime models have offered styles such as Woodblock print and Ink wash for a while. Under the hood, though, those presets only added a few style words to your prompt. When we lined up the same prompt with and without them, the picture barely changed: it was still a clean anime illustration with a little texture on top.

So we trained small style models of our own, one per style, and connected them to four presets: Woodblock print (ukiyo-e), Ink wash, Art Nouveau and a new Impressionism preset.

Trained only on public-domain art

We did not use images from the web at large, and we did not use anything made on EGAKU. Each style model learned from about 700 prints or paintings that are in the public domain or released under CC0, chosen from an open collection that only contains such works.

  • No single artist dominates. We capped each artist at 120 works, so the woodblock style is not one painter's style.
  • Names were removed. We took artist names out of the descriptions the model learns from, so it learns the look, not a name to imitate.
  • People in the training images are adults only. We excluded every image whose description mentioned a child, and every image with nudity.

Before and after

Same prompt, same seed, only the style model added.

Woodblock print: indigo skies, vermilion railings, kimono patterns and the flat printed colour of real prints.

Two people on a red bridge in ukiyo-e style — Before: style tags only
Before: style tags only
Two people on a red bridge in ukiyo-e style — Now: our own style model
Now: our own style model
A samurai under a cherry tree in ukiyo-e style — Before: style tags only
Before: style tags only
A samurai under a cherry tree in ukiyo-e style — Now: our own style model
Now: our own style model

Ink wash: bleeding ink, empty space and paper tone instead of solid black and white.

A cat on a tiled roof in ink wash style — Before: style tags only
Before: style tags only
A cat on a tiled roof in ink wash style — Now: our own style model
Now: our own style model

Art Nouveau: the old preset already drew an ornamental frame; the new model adds the colour and paper texture of the posters.

Two people holding hands in Art Nouveau style — Before: style tags only
Before: style tags only
Two people holding hands in Art Nouveau style — Now: our own style model
Now: our own style model

Impressionism (new): dappled light, visible brush strokes and soft edges.

A city street at night in Impressionist style — Before: style tags only
Before: style tags only
A city street at night in Impressionist style — Now: our own style model
Now: our own style model

Stronger was not better

Turned up high, the style models start to overrule your prompt. In the ink wash test, two people on a bridge became one. The woodblock model, which saw many kabuki actor prints, turned a woman adventurer into a male actor's face.

Two people on a bridge in ink wash style — Too strong: one person is lost
Too strong: one person is lost
Two people on a bridge in ink wash style — Tuned: both people stay
Tuned: both people stay

For each style we tried four strengths on five different subjects and chose the strongest setting that still kept what was asked for: the right number of people, the right subject. One limit remains: in the woodblock style, a cat still tends to turn into a person, because there are almost no cats in the prints it learned from.

Fewer extra arms and legs

Two other changes went out at the same time, aimed at the most common complaint about AI images: an extra arm, a leg in the wrong place, or two bodies blending together when people hug or dance.

  • Pick the clean one. When your prompt has two people together, a hug, or a big movement such as a jump or a kick, our own image models now make one extra candidate. An automatic check looks at each one for extra or misplaced limbs, merged bodies and the wrong number of people, and you get the clean one. You pay for the images you asked for, not the extra one.
  • A cleaner finishing pass for α2. Our photo-style model α2 enlarges each image and redraws the details. We changed how it enlarges before that redraw. In our test of 24 hard poses, the problem cases went from 8 to 6, the faint floating fragments around bodies disappeared, and each image finishes about a third faster.
Close-up of a dancing couple — Old finishing pass
Old finishing pass
Close-up of a dancing couple — New finishing pass
New finishing pass

None of this makes mistakes impossible, but it makes them much rarer. Next we are working on fixing the pose first for acrobatic and multi-person scenes, so the body has a correct shape before any detail is drawn.

Essayez par vous-même

Créez-le avec l’outil présenté dans cet article. Crédits offerts à l’inscription.

Ouvrir l’outil, c’est gratuit

Plus d’articles

how-to

Flyers et enseignes par IA avec du chinois et du coréen bien écrits

5 min de lecture · 2026-11-20

how-to

Préparez une scène en 3D et obtenez cinq angles de caméra à partir d’un seul décor

6 min de lecture · 2026-11-20

how-to

Placer la photo de votre produit dans une nouvelle scène avec l’IA (exemples, prompts et ce qui n’a pas marché)

6 min de lecture · 2026-11-18

how-to

Faire un flyer avec l’IA sans qu’il fasse « IA »

5 min de lecture · 2026-11-15