跪拜 Guibai
← All articles
Developer · Artificial Intelligence · AIGC

A Single Prompt Turns a Photo Into a Photoreal-and-Anime Dual Portrait

By 怕浪猫 ·
Read original on juejin.cn ↗ Google Translate ↗ Alt translation

The prompt demonstrates how far a single, well-constrained text instruction can go in a modern image model — producing a specific multi-character composition with consistent likeness, distinct art styles, and controlled emotional tone, all without post-processing or inpainting.

Summary

The prompt uses an uploaded photo as a definitive likeness reference and produces a tall, high-end composition. One side renders the subject photorealistically; the other places a smaller anime version of the same person in an affectionate pose — clinging to an arm, making finger hearts, or gazing upward with bright eyes. The real-life figure wears a mock-grumpy expression that reads as fond rather than hostile.

A soft pastel backdrop and flattering light push the aesthetic toward Korean influencer content. The prompt explicitly forbids anatomical errors, extra limbs, visual clutter, text, watermarks, and any tone that feels cold or unsettling. The result is a polished image built from a single, carefully constrained generation request.

Takeaways
A single prompt can generate a split portrait containing both a photorealistic subject and a stylized anime version of the same person.
The anime figure is directed to perform specific affectionate actions such as clinging, making finger hearts, or gazing upward.
The real-life figure is instructed to show a mock-grumpy expression that still conveys hidden fondness.
Negative prompts block anatomical errors, extra limbs, 3D-doll appearance, text, watermarks, and unsettling tones.
The aesthetic targets a soft pastel Korean influencer look with flattering light and polished quality.
Conclusions

The prompt's structure treats the anime clone not as a separate subject but as an extension of the same identity, which reduces the risk of mismatched facial features.

Specifying an emotional dynamic between the two figures — overt affection versus playful annoyance — gives the model a narrative constraint that likely improves coherence across the composition.

Negative prompts here do more than suppress artifacts; they define the emotional floor of the image by forbidding hostility, coldness, and unsettling vibes.

Source: juejin.cn ↗ Google Translate ↗ Backup ↗