[S.021] Starryear Milk Frog
Turn photos into surreal giant milk frogs
[S.021] Starryear Milk Frog
Turn photos into surreal giant milk frogs
Showcase
Description
Turn a photo into a unique surreal “milk frog avatar” scene, making the people, animals, or key objects in the image reappear as giant milk frogs with transformed poses, scale, and spatial relationships. The milk frog retains its rounded head, frog eyes, large belly, creamy belly markings, and smooth, soft-clay-like texture. At the same time, the image’s distinctive elements are naturally integrated into the head, body, or limbs, creating absurd forms that look as if they grew from the character itself rather than being simple decorations. By default, the result is a 9:16 vertical diptych: the complete original photo remains at the top as a visual reference, while the bottom presents a reconstructed scene dominated by the milk frog. The image carries forward recognizable poses, silhouettes, colors, decorations, and environmental cues from the original, while exaggerated scale creates a monumental visual impact. Original materials remain primarily in the environment and scene, while the milk frog itself retains a soft, rounded, clean surface. This is ideal for transforming travel photos, personal styles, statues, buildings, plants, or everyday objects into absurd yet clear visual works.
Recommended by
淡苍
Why we love this skill
Turns any photo into a surreal vertical diptych with original-image evidence and a giant milk-frog avatar, naturally integrating key elements while preserving visual cues. Stable character anchors and photo-driven motion balance absurdity with visual storytelling.
Best for
For creators who want to turn travel photos, personal styles, or everyday objects into absurd, monumental visual works.
How to use this Skill
Provide your input
Upload a photo featuring a person, animal, building, plant, or everyday object.
Run the Skill
Transform the subject and key visual cues into a giant milk frog avatar while keeping the original photo as the compositional reference.
Review your result
Generate a 9:16 vertical surreal diptych: the original photo appears above, while a giant milk frog scene incorporating its elements appears below.
Related Skills
View all
ImageS.010 · Starryear Absurdist
Turn one photo into a vertical absurd triptych that uses the original image as visual evidence: the top panel presents a scale-shifted absurd scene, the middle preserves the original photo with a horizontal crop, and the bottom forms a return or causal loop from the same absurd relationship. The final image combines three equal-width landscape 16:9 panels vertically with no gaps, creating an exact 16:27 composition without borders, text, or watermarks. The generated top and bottom panels develop new relationships only from the people, animals, objects, actions, quantities, colors, and lighting in the photo. They preserve important identifying features and recognizable outlines, using a spacious, restrained real-world or puppet-theater aesthetic. The middle panel uses only the original photo pixels, with orientation correction, proportional scaling, and conservative cropping where necessary. It is not redrawn, retouched, or filled in, ensuring that it remains a reliable record of the original image. Ideal for vertical absurd triptychs with the original image in the middle, photo-based derivative art, and surreal posters with visual narratives. It is not intended for ordinary filters, unrelated fantasy scenes, dense collages, or layouts that redraw the middle evidence photo.
Image【S.023】Starryear · 3D Photo
Turn a real photo into a vertical “memory nine-grid” artwork that combines documentary grounding with a surreal sense of space. The upper section preserves a centered square evidence panel from the original photo, while the lower section expands into a complete 3D nine-grid world. People, objects, movements, color relationships, and environmental cues from the photo are rearranged across the nine panels to create a coherent spatial paradox with a photographic feel—not an ordinary filter or simple crop-and-paste collage. The main subject is enlarged in the lower grid as the primary visual focus, clearly preserving the person’s face, pose, or key object while crossing adjacent panels and dividing lines. Foreground and background occlusion, perspective, depth, and natural shadows create a convincing sense of three-dimensional layering. Each panel retains a surreal transformation driven by the source photo, while all nine share the same spatial logic, allowing fine details, expansive views, and negative space to form one cohesive scene. The final output is a vertical PNG with no text or watermark. The upper and lower sections are unified through coordinated white balance, color temperature, brightness, contrast, and saturation, allowing the original photo and reconstructed scene below to blend naturally at the transition. Suitable for surreal photography, photo memory art, visual storytelling with people, and creative images with collectible appeal.
Image[S.012] Year Clone · Photo
Turn one photo into a borderless vertical triptych: the center preserves an unaltered crop of the original image as clear evidence of the scene, while the top and bottom each clone and invert a different subject from the photo. The result is a surreal image rooted in the original environment. The top panel can focus on a primary subject such as a person or animal, while the bottom panel works well with boats, buildings, railings, landmarks, or other clearly structured objects. Each generated area retains one recognizable subject at a clear scale, while allowing smaller inverted forms to blend into the foreground, background, water, rocks, roads, or building edges. The original scene’s lighting, perspective, occlusion, materials, and depth of field are maintained. Ideal for photographic collages with separately duplicated and inverted people and objects, source-photo-driven surreal triptychs, and visual experiments that preserve features such as landmarks, mountains, and skylines. The final output is a single vertical 2:3 image with no text, border, watermark, or decorative title.
Information
- Version
- v6
- Last updated
- Runtime credits
- Usage-based
- Models
- Auto
- Use cases
- Images & designLifestyle & entertainment
- What you get
- Image
- Visual styles
- 3DSurrealist