# Interpretation of prompt words

What are prompts? **A prompt is a specialized instructional language for image generation that guides AI systems to create desired images through precise textual descriptions.** Imagine you're talking to a genius artist who only understands written language. How would you accurately describe the image in your mind to them? This is the magic of "prompts" in AI art generation.

How can we write precise prompts for ReadKidz? Here's a detailed interpretation of ReadKidz prompts! **Basic Prompt Format** A complete ReadKidz prompt typically includes the following parts: ① Single-person scene prompt format : \[Scene shooting method]\[solo]\[Time-space description]\[Character shooting method]\[Character information] ② Two-person scene prompt format : \[Number of people]\[Spatial composition]\[Scene shooting method]\[Time-space description]\[Character interaction words]\[Character 1 shooting method]\[Character 1 information]\[Character 2 shooting method]\[Character 2 information] ③ Empty scene prompt format : \[Scene shooting method]\[Time-space description] **1. Number of People in the Scene** The number of people in the scene directly affects the distribution of focus and overall atmosphere in the image. We need to clearly **define the number of characters** that will appear in the frame. For example : For a single-person scene, use "solo"; for a two-person scene, use "Two-character scene"; for scenes without people, use "Empty scene";

**2. Spatial Composition** The composition method determines the organization and arrangement of visual elements, directly influencing the viewer's visual guidance and areas of focus. Appropriate composition can significantly enhance the visual appeal and artistic expression of the image. We need to **describe the spatial layout and organization of the image**, including composition types, aspect ratios, and visual balance methods. For example : symmetrical composition, rule of thirds, golden ratio, central symmetry, open composition, etc.

**3. Scene Shooting Method** The shooting method greatly influences the visual style and sense of space in the image, creating unique artistic effects. Different "shooting" techniques bring various spatial compression or expansion effects, affecting the dramatic impact and emotional expression of the image. We need to **describe the presentation technique, similar to camera shots and technical treatments** in photography or film. For example : wide-angle shot, medium shot, close-up shot, etc.

**4. Time and Space Description** The time and space description establishes the basic context for the image, determining lighting, atmosphere, and environmental elements. This component directly affects the tone, lighting effects, and overall mood, serving as a key part of the image's narrative. We need to **specify the temporal background** (historical period, time of day, etc.) and **establish the location and environment**, which may include weather conditions, seasons, etc. For example : dusk, misty morning, futuristic city, medieval castle, etc.

**5. Character Shooting Method** The character shooting method determines how to focus on and present character features, which is crucial for expressing emotions and personality. Different angles and distances can convey varying emotional impacts and levels of intimacy. We need to **describe the angle and way characters are presented in the image**, similar to techniques used in portrait photography. For example : close-up shot, half-length portrait, full-body shot, side profile, etc.

**6. Character Information** Character information is central to the visual presentation of people, directly determining the main image and narrative focus. Detailed character descriptions help AI generate expected character images, enhancing storytelling and emotional expression. We need to **describe the character's appearance, clothing, expressions, and poses**, including characteristics such as age, occupation, emotional state, as well as behaviors and interactions. For example : a young woman in a red dress with flowing hair; an elderly fisherman with a weather-beaten face in blue workwear; a little girl in a blue T-shirt and shorts, swinging on a swing.

**Prompt Optimization Tips** **1. Following Structural Order Principles** Structured prompts help AI process information in a logical sequence, first determining scene scale, then establishing composition framework, followed by time-space background, and finally focusing on character details. We need to follow the natural order from scene to details, allowing AI to build the image progressively. Optimization Example : Basic prompt: magical forest, little girl, mushroom, daytime Optimized prompt : solo scene, central composition, macro shot, misty morning magical forest, close-up shot, red-haired little girl holding a glowing mushroom

**2. Precise Descriptions Instead of Vague Adjectives** AI better understands specific, visualizable descriptions rather than subjective evaluative terms. "Beautiful" and "cute" are subjective concepts, less effective than direct descriptions of specific features. We need to replace general adjectives with specific, precise descriptions. Optimization Example : Basic prompt : beautiful forest, cute little girl Optimized prompt : solo scene, rule of thirds composition, bird's eye view, magical forest with sunlight filtering through leaves, close-up shot, red-haired little girl in a green dress

**3. Controlling Information Density in Each Section** Too much redundant or contradictory information confuses AI, leading to cluttered image elements. Emphasize core elements while keeping supporting elements concise. We need to provide more details for the most important parts while appropriately simplifying secondary elements. Optimization Example : Unbalanced : solo scene, overly complex rule of thirds with diagonal mixed composition, fisheye wide-angle macro mixed shooting, misty morning magical forest, close-up shot, little girl Optimized : solo scene, rule of thirds composition, macro shot, misty morning magical forest, close-up shot, red-haired little girl picking glowing mushrooms

**4. Using Contrast to Emphasize Focus** Contrasts in light and shadow, color, and size can guide visual focus and make the subject more prominent. We need to enhance the visual effect of the main subject through contrasting elements. Optimization Example : Basic prompt : solo scene, central composition, magical forest, little girl Optimized prompt : solo scene, central composition, macro shot, against a dark magical forest background, close-up shot, little girl in bright red dress holding a glowing mushroom

**5. Linking Emotional Atmosphere with Visual Elements** Emotional atmosphere and visual elements should support each other to create a coherent visual experience. We need to ensure emotional descriptions match visual elements. Optimization Example : Inconsistent prompt : solo scene, bright magical forest, horror atmosphere, little girl picking mushrooms Optimized prompt : solo scene, central composition, bright magical forest, cheerful atmosphere, close-up shot, little girl with bright smile discovering giant colorful mushrooms

**Prompt Challenges** **1. Character Image Stability Control** AI struggles to maintain consistent character appearance across multiple images, especially facial features, clothing details, and body proportions. We can create a unique and repeatable character description by specifying details like age, facial features, hairstyle, clothing color, and style. These specific characteristics enable AI to generate consistent character images across multiple pictures, rather than reinterpreting the vague concept of "little girl" each time. Optimization Example : Basic prompt 1: solo scene, central composition, garden scene, close-up shot, little girl in garden Basic prompt 2: solo scene, central composition, garden scene, close-up shot, little girl picking flowers Optimized prompt 1: solo scene, central composition, garden scene, close-up shot, 6-year-old girl Lily, round face, big eyes, brown curly hair in pigtails, blue overall dress, white T-shirt, in garden Optimized prompt 2 : solo scene, central composition, garden scene, close-up shot, 6-year-old girl Lily, round face, big eyes, brown curly hair in pigtails, blue overall dress, white T-shirt, picking flowers

**2. Challenges in Two-Person Scene Interaction Composition** In two-person scenes, AI often struggles to accurately represent character interactions, relative positions, and proportional relationships. We need to clearly specify the spatial relationship between two characters (left/right) and describe specific interactive behaviors (handing balloon) rather than simple juxtaposition. This modification helps AI understand the relationship between characters and the flow of action, creating meaningful interactive scenes rather than merely placing two independent characters in the same frame. Optimization Example : Basic prompt: two-person scene, rule of thirds composition, medium shot, park, close-up shot, boy and girl in park, playing with balloon Optimized prompt : two-person scene, rule of thirds composition, medium shot, park, close-up shot, 8-year-old boy Tom with short black hair and blue T-shirt standing on the left handing balloon to 6-year-old girl Sarah with blonde ponytail and pink dress standing on the right

After you have fully understood the basic format and optimization techniques of prompt words, let's write prompt words as a whole! **Single scene prompt words** : Mid-shot, side view, morning, park path, soft morning light, stone path beside the flower bed, morning sun casting soft shadows, park bench in the background, flowers blooming all around, contemplative and warm atmosphere, solo, portrait, side view, little boy in blue T-shirt, child, contemplative, standing thoughtfully, head slightly tilted, white skin, brown hair, short hair, straight hair, medium-length eyebrows, curved eyebrows, big eyes, brown eyes, round eyes, small nose, round ears, medium-sized ears, small mouth, upturned corners of the mouth, round face, plump cheeks, soft jawline, blue T-shirt, brown overalls, white sneakers

**Two-person scene prompts** : Two-person scene, diagonal composition, simple tree-lined avenue, wide shot, side view, eye-level view, afternoon, on a park path, flowing sunlight, energetic atmosphere, boy and dog running together in harmony, Jack (child, taller in stature, medium shot, side view, eye-level view, running with a determined pace, focused expression, white skin, short straight brown hair, large round brown eyes, button nose and small upturned mouth, blue t-shirt, brown overalls and white sneakers), Rex (golden retriever puppy, shorter in stature, medium shot, side view, eye-level view, trotting together with a matching pace, alert expression, golden fur, brown eyes, black nose and floppy ears)

**Aerial scene prompts** : Wide-angle lens, side view, bird's-eye view, morning, park, soft natural light, lush green grass, blooming flowers, blue sky and scattered white clouds, paved paths winding through the landscape, park benches scattered everywhere, warm and peaceful atmosphere, breeze blowing the leaves, fresh floral fragrance floating in the air Of course, this is only the prompts written for a single illustration. In order to ensure the stability of the characters and pictures in the picture book, ReadKidz will select the picture book style and the raw pictures of the character portraits before the picture book subtitles are drawn.

Optimization Creation Tips : Ensuring Consistency and Stability of ReadKidz Picture Books When creating picture books with ReadKidz, the system automatically generates corresponding prompts. However, to create children's picture books with excellent consistency and stability, we need to pay special attention to the following key points : **1. Character Image Consistency Maintenance** The descriptive prompts on the character portrait generation page must precisely match the final generated character portraits. This includes not only **basic features** (hair color, clothing, etc.) but also **facial features**, **body proportions**, and **unique identifiers**. When you see that the generated character portrait doesn't match your expectations, you should immediately **adjust the prompts**.

**2. Character Image Selection on Storyboard Page** The character images provided on the storyboard page are key tools for maintaining character image stability. When reviewing prompts, you must carefully **compare the prompt content and select the corresponding character images**.

**3. Handling Complex Two-Person Scenes** For two-person scenes that are difficult to generate (such as complex interactive poses, unnatural proportional relationships), we can take strategic approaches : ① **Break down complex two-person scenes into single-person scenes** while maintaining narrative flow ② Simplify character interactions, **focusing on basic interactions** that are easier to generate (such as face-to-face conversations, standing side by side) ③ **Use environmental elements** (such as tables, trees) to naturally separate characters, reducing the complexity of direct interaction For example : An anthropomorphic deer watching an anthropomorphic rabbit emerging from the bushes!

By strictly following these optimization suggestions, you will be able to significantly improve the visual consistency and character stability of ReadKidz picture books, creating professional-quality children's picture book works. --- # Agent Instructions: Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter: ``` GET https://readkidz.gitbook.io/readkidz/readkidz-getting-started-tutorial/interpretation-of-prompt-words.md?ask= ``` The question should be specific, self-contained, and written in natural language. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.