Models>

Marvel promo cards 2023 (style/template)

Quick Generate
Select a style, write what you want to see and click Generate
What you want in the image.
Clear
Random
* This model is not available for use and might generate different characters compared to the ones seen in the images.

Model Description by Creator

Pack of LyCORIS/LoCon models to replicate the style and hud shown on Marvel promotional cards given in E. Leclerc stores during oct. - dec. 2023.
Be creative, give your favorite characters this 'super' touch !

I did my best to remove scanning artifacts but it still has flaws, it might show some vertical color stripes at high weights. Sorry !
Note: french texts. Puissance == power

Cooking recipe:
- Scan every single of them. I don't have the full deck but almost. See below for details.
- Descreen/upscale: this model did great, althrough not perfect: https://openmodeldb.info/models/2x-BIGOLDIES
- Deskew: semi auto with https://github.com/galfar/deskew/releases/tag/v1.30
- Reshape, cut round corners, dust removal, color/contrast fix: manual, half PS, half Gimp (very long...)
- For each of them, find the positions for the stat wheel(s), then generate all possible combinations: 11 for a hero card, 11^3 = 1331 (!!) for villain, 8 for a powerup. Total of close to 20k images...
- Build controlnet templates. Fonts are not a perfect match but those are the closest i could find:
Character name -> Annapolis: https://www.dafont.com/annapolis.font
Index -> Denmark: https://fontzone.net/font-details/denmark-regular
(both free)
- Controlnet models for SDXL:
https://huggingface.co/lllyasviel/sd_control_collection/tree/main
- I tried to train them as a pack with different triggers but it didn't work well, so i had to split hero/villain/powerup. Bit annoying but no big deal.
- I haven't found a way to constraint colors on the hud, sometimes it's good sometimes... bit less.
- Index numbers is not very stable even with controlnet and may require inpainting to cherry pick it.

Weights:
- Hero: 0.7-0.9
Controlnet required for character name, HUD is a hit and miss without it.
- Villain: 0.8
Controlnet mandatory for character name and the HUD, or it generates a ghost 4th stat wheel in the middle.
- Powerup: 0.8-0.9
Controlnet required only for the stat name, HUD is fine by itself.

Base model:
- Anything that can draw comics-style stuff. Samples are done with Cheyenne.

Triggers & keywords:
- Trigger marvelpromocard2023 is used on all of the models. Then each has its specifics:
- Hero main trigger is herocard, with one extra keyword among intelligenceN, powerN, courageN, where N is in 0 - 10 (inclusive)
- Villain main trigger is villaincard, with 3 extra keywords as intelligenceN, powerM, courageP where N, M, P are in 0 - 10 (inclusive)
- Power-up main trigger is vpowerupcard as they are only designed to power up villains, with one extra keyword among intelligenceN, powerN, courageN, where N is in 1 - 8 (inclusive). Dataset often features objects or hand-held weapons and will always show some shining grey lighting streaks behind stuff.
- On top of that you can specify index in the form indexXXX which is going to be the card number in the deck, from 001 to 099 to the exception of 002, 011 and 062 which are the ones i don't have (damn rares...). Old school Avengers cards at index 100+ from the original collection aren't included in the datasets.

Prompts & usage:

Size: Best with XL 832x1216 (2:3) format, 50+ steps, DPM++ 2M/3M Karras

Standard prompt template:

(best quality, masterpiece, absurdres, highres, ultra_detailed, dynamic angle:1.2), herocard, intelligence10, (your subject/background here), index034, (intricate details, hyperdetailed:1.15), (ultrahigh resolution textures), bokeh, (volumetric, cinematic) lighting, depth of field, <lora:marvel_promo_cards_2023_locon_xl_hero_v1:0.9>

Also remove text from the negative and add inaccurate text, doubled letters.

See samples for examples.

Extra shine with controlnet:

Use ControlNet/Canny with diffusers_xl_full an input images generated from one of the templates.
To keep the rounded corners enable the 'Black matte' layer. For a squared image, disable it.

Have fun and may the best win !

Images Generated With This Model

(best quality, masterpiece, absurdres, highres, ultra_detailed, dynamic angle:1.2), herocard, intelligence10, portrait of the dalai lama, glasses, buddhist temple background, index034, (intricate details, hyperdetailed:1.15), (ultrahigh resolution textures), bokeh, (volumetric, cinematic) lighting, depth of field, <lora:marvel_promo_cards_2023_locon_xl_hero_v1:0.8>
N/A
Hidden For Public Users
(best quality, masterpiece, absurdres, highres, ultra_detailed, dynamic angle:1.2), herocard, intelligence10, portrait of the dalai lama, glasses, buddhist temple background, index034, (intricate details, hyperdetailed:1.15), (ultrahigh resolution textures), bokeh, (volumetric, cinematic) lighting, depth of field, <lora:marvel_promo_cards_2023_locon_xl_hero_v1:0.8>
(best quality, masterpiece, absurdres, highres, ultra_detailed, dynamic angle:1.2), herocard, intelligence6, portrait of Morticia Addams, inside an old derelict manor, index023, (intricate details, hyperdetailed:1.15), (ultrahigh resolution textures), bokeh, (volumetric, cinematic) lighting, depth of field, <lora:marvel_promo_cards_2023_locon_xl_hero_v1:0.8>
N/A
Hidden For Public Users
(best quality, masterpiece, absurdres, highres, ultra_detailed, dynamic angle:1.2), herocard, intelligence6, portrait of Morticia Addams, inside an old derelict manor, index023, (intricate details, hyperdetailed:1.15), (ultrahigh resolution textures), bokeh, (volumetric, cinematic) lighting, depth of field, <lora:marvel_promo_cards_2023_locon_xl_hero_v1:0.8>
(best quality, masterpiece, absurdres, highres, ultra_detailed, dynamic angle:1.2), herocard, courage9, portrait of Batman, black long cape, cityscape, clock tower, by night, view from below, contrapposto, index012, (intricate details, hyperdetailed:1.15), (ultrahigh resolution textures), bokeh, (volumetric, cinematic) lighting, depth of field, <lora:marvel_promo_cards_2023_locon_xl_hero_v1:0.8>
N/A
Hidden For Public Users
(best quality, masterpiece, absurdres, highres, ultra_detailed, dynamic angle:1.2), herocard, courage9, portrait of Batman, black long cape, cityscape, clock tower, by night, view from below, contrapposto, index012, (intricate details, hyperdetailed:1.15), (ultrahigh resolution textures), bokeh, (volumetric, cinematic) lighting, depth of field, <lora:marvel_promo_cards_2023_locon_xl_hero_v1:0.8>
(best quality, masterpiece, absurdres, highres, ultra_detailed, dynamic angle:1.2), herocard, power8, portrait of Duke Nukem, holding big guns firing, inside a space station, index035, indfuturistic, sci-fi, (intricate details, hyperdetailed:1.15), (ultrahigh resolution textures), bokeh, (volumetric, cinematic) lighting, depth of field, <lora:marvel_promo_cards_2023_locon_xl_hero_v1:0.8>
N/A
Hidden For Public Users
(best quality, masterpiece, absurdres, highres, ultra_detailed, dynamic angle:1.2), herocard, power8, portrait of Duke Nukem, holding big guns firing, inside a space station, index035, indfuturistic, sci-fi, (intricate details, hyperdetailed:1.15), (ultrahigh resolution textures), bokeh, (volumetric, cinematic) lighting, depth of field, <lora:marvel_promo_cards_2023_locon_xl_hero_v1:0.8>
(best quality, masterpiece, absurdres, highres, ultra_detailed, dynamic angle:1.2), herocard, courage9, Indiana Jones, (1boy:1.3), holding an ancient aztec relic, inside an ancient temple full of traps, index081, (intricate details, hyperdetailed:1.15), (ultrahigh resolution textures), bokeh, (volumetric, cinematic) lighting, depth of field, <lora:marvel_promo_cards_2023_locon_xl_hero_v1:0.8>
N/A
Hidden For Public Users
(best quality, masterpiece, absurdres, highres, ultra_detailed, dynamic angle:1.2), herocard, courage9, Indiana Jones, (1boy:1.3), holding an ancient aztec relic, inside an ancient temple full of traps, index081, (intricate details, hyperdetailed:1.15), (ultrahigh resolution textures), bokeh, (volumetric, cinematic) lighting, depth of field, <lora:marvel_promo_cards_2023_locon_xl_hero_v1:0.8>
(best quality, masterpiece, absurdres, highres, ultra_detailed, dynamic angle:1.2), herocard, courage6, Sonic the hedgehog, running, 2D arcade game, 1990s game, index003, (intricate details, hyperdetailed:1.15), (ultrahigh resolution textures), bokeh, (volumetric, cinematic) lighting, depth of field, <lora:marvel_promo_cards_2023_locon_xl_hero_v1:0.8>
N/A
Hidden For Public Users
(best quality, masterpiece, absurdres, highres, ultra_detailed, dynamic angle:1.2), herocard, courage6, Sonic the hedgehog, running, 2D arcade game, 1990s game, index003, (intricate details, hyperdetailed:1.15), (ultrahigh resolution textures), bokeh, (volumetric, cinematic) lighting, depth of field, <lora:marvel_promo_cards_2023_locon_xl_hero_v1:0.8>
(best quality, masterpiece, absurdres, highres, ultra_detailed, dynamic angle:1.2), herocard, courage7, portrait of Obi-Wan Kenobi, Tatooine desert, index017, (intricate details, hyperdetailed:1.15), (ultrahigh resolution textures), bokeh, (volumetric, cinematic) lighting, depth of field, <lora:marvel_promo_cards_2023_locon_xl_hero_v1:0.8>
N/A
Hidden For Public Users
(best quality, masterpiece, absurdres, highres, ultra_detailed, dynamic angle:1.2), herocard, courage7, portrait of Obi-Wan Kenobi, Tatooine desert, index017, (intricate details, hyperdetailed:1.15), (ultrahigh resolution textures), bokeh, (volumetric, cinematic) lighting, depth of field, <lora:marvel_promo_cards_2023_locon_xl_hero_v1:0.8>
(best quality, masterpiece, absurdres, highres, ultra_detailed, dynamic angle:1.2), herocard, intelligence7, portrait of Wednesday Addams (dressed as a apache girl, feathered clothes, native american headdress:1.3), light smile, (holding a lit match:1.3), outdoor, forest, index024, (intricate details, hyperdetailed:1.15), (ultrahigh resolution textures), bokeh, (volumetric, cinematic) lighting, depth of field, <lora:marvel_promo_cards_2023_locon_xl_hero_v1:0.8>
N/A
Hidden For Public Users
(best quality, masterpiece, absurdres, highres, ultra_detailed, dynamic angle:1.2), herocard, intelligence7, portrait of Wednesday Addams (dressed as a apache girl, feathered clothes, native american headdress:1.3), light smile, (holding a lit match:1.3), outdoor, forest, index024, (intricate details, hyperdetailed:1.15), (ultrahigh resolution textures), bokeh, (volumetric, cinematic) lighting, depth of field, <lora:marvel_promo_cards_2023_locon_xl_hero_v1:0.8>
(best quality, masterpiece, absurdres, highres, ultra_detailed, dynamic angle:1.2), herocard, intelligence8, portrait of Shuri in a black panther suit with purple glowing bits, index099, (intricate details, hyperdetailed:1.15), (ultrahigh resolution textures), bokeh, (volumetric, cinematic) lighting, depth of field, <lora:marvel_promo_cards_2023_locon_xl_hero_v1:0.8>
N/A
Hidden For Public Users
(best quality, masterpiece, absurdres, highres, ultra_detailed, dynamic angle:1.2), herocard, intelligence8, portrait of Shuri in a black panther suit with purple glowing bits, index099, (intricate details, hyperdetailed:1.15), (ultrahigh resolution textures), bokeh, (volumetric, cinematic) lighting, depth of field, <lora:marvel_promo_cards_2023_locon_xl_hero_v1:0.8>
(best quality, masterpiece, absurdres, highres, ultra_detailed, dynamic angle:1.2), vpowerupcard, intelligence4, a strange green glowing (television:1.4) capable of controlling minds, no human, index007, futuristic, sci-fi, (intricate details, hyperdetailed:1.15), (ultrahigh resolution textures), bokeh, (volumetric, cinematic) lighting, depth of field, <lora:marvel_promo_cards_2023_locon_xl_vpowerup_v1:0.8>
N/A
Hidden For Public Users
(best quality, masterpiece, absurdres, highres, ultra_detailed, dynamic angle:1.2), vpowerupcard, intelligence4, a strange green glowing (television:1.4) capable of controlling minds, no human, index007, futuristic, sci-fi, (intricate details, hyperdetailed:1.15), (ultrahigh resolution textures), bokeh, (volumetric, cinematic) lighting, depth of field, <lora:marvel_promo_cards_2023_locon_xl_vpowerup_v1:0.8>
(best quality, masterpiece, absurdres, highres, ultra_detailed, dynamic angle:1.2), vpowerupcard, power5, a sith red lightsaber with red lightning streaks, no human, futuristic, sci-fi, (intricate details, hyperdetailed:1.15), (ultrahigh resolution textures), bokeh, (volumetric, cinematic) lighting, depth of field, <lora:marvel_promo_cards_2023_locon_xl_vpowerup_v1:0.8>
N/A
Hidden For Public Users
(best quality, masterpiece, absurdres, highres, ultra_detailed, dynamic angle:1.2), vpowerupcard, power5, a sith red lightsaber with red lightning streaks, no human, futuristic, sci-fi, (intricate details, hyperdetailed:1.15), (ultrahigh resolution textures), bokeh, (volumetric, cinematic) lighting, depth of field, <lora:marvel_promo_cards_2023_locon_xl_vpowerup_v1:0.8>
(best quality, masterpiece, absurdres, highres, ultra_detailed, dynamic angle:1.2), villaincard, intelligence3, power7, courage1, portrait of Donald Trump, business suit, speaking on a podium, american flags, (red and black hud with white numbers), (intricate details, hyperdetailed:1.15), (ultrahigh resolution textures), bokeh, (volumetric, cinematic) lighting, depth of field, <lora:marvel_promo_cards_2023_locon_xl_villain_v1:0.8>
N/A
Hidden For Public Users
(best quality, masterpiece, absurdres, highres, ultra_detailed, dynamic angle:1.2), villaincard, intelligence3, power7, courage1, portrait of Donald Trump, business suit, speaking on a podium, american flags, (red and black hud with white numbers), (intricate details, hyperdetailed:1.15), (ultrahigh resolution textures), bokeh, (volumetric, cinematic) lighting, depth of field, <lora:marvel_promo_cards_2023_locon_xl_villain_v1:0.8>

Train Models

Train AI Models On Your Own Pictures

You can use your own pictures to train AI models. Reimagine yourself in new ways with our AI-powered image generator.
Create pictures of yourself around the world, in your dream outfit, as a video game character, or in countless art styles. The possibilities are endless.
Train Your Own Model

Generate AI Images. Simple and Easy

AIEasyPic is the #1 AI image generator. Generate images with simple prompts, swap faces, train models to reimagine yourself, and more.