Character Design Sheet

角色设计参考表

Create consistent characters across multiple AI-generated images via inference.sh CLI.

通过inference.sh CLI在多张AI生成图像中创建一致性角色。

Quick Start

快速开始

bash

curl -fsSL https://cli.inference.sh | sh && infsh login

bash

curl -fsSL https://cli.inference.sh | sh && infsh login

Generate a character concept

生成角色概念

infsh app run falai/flux-dev-lora --input '{ "prompt": "character design reference sheet, front view of a young woman with short red hair, green eyes, wearing a blue jacket and white t-shirt, full body, white background, clean lines, concept art style, character turnaround", "width": 1024, "height": 1024 }'

undefined

infsh app run falai/flux-dev-lora --input '{ "prompt": "character design reference sheet, front view of a young woman with short red hair, green eyes, wearing a blue jacket and white t-shirt, full body, white background, clean lines, concept art style, character turnaround", "width": 1024, "height": 1024 }'

undefined

The Consistency Problem

一致性问题

AI image generation produces different-looking characters every time, even with the same prompt. This is the #1 challenge in AI art for any project requiring the same character across multiple images.

即使使用相同提示词，AI图像生成每次都会生成外观不同的角色。这是所有需要在多张图像中使用同一角色的AI美术项目面临的头号挑战。

Solutions (Ranked by Effectiveness)

解决方案（按有效性排序）

Technique	Consistency	Effort	Best For
FLUX LoRA (trained on character)	Very high	High (requires training data)	Ongoing projects, many images
Detailed description anchor	Medium-high	Low	Quick projects, few images
Same seed + similar prompt	Medium	Low	Variations of single pose
Image-to-image refinement	Medium	Medium	Refining existing images
Reference image in prompt	Varies	Low	When model supports it

技术	一致性	投入成本	适用场景
FLUX LoRA（基于角色训练）	极高	高（需要训练数据）	长期项目、大量图像生成
详细描述锚点	中高	低	快速项目、少量图像生成
相同种子+相似提示词	中等	低	单一姿势的变体生成
图生图优化	中等	中等	优化现有图像
提示词中加入参考图像	不稳定	低	模型支持该功能时

Reference Sheet Types

参考表类型

1. Turnaround Sheet

1. 多角度视图表

Shows the character from multiple angles:

┌────────┬────────┬────────┬────────┐
│        │        │        │        │
│ FRONT  │  3/4   │  SIDE  │  BACK  │
│  VIEW  │  VIEW  │  VIEW  │  VIEW  │
│        │        │        │        │
└────────┴────────┴────────┴────────┘

bash

undefined

展示角色的多个角度：

┌────────┬────────┬────────┬────────┐
│        │        │        │        │
│ 正面   │ 四分之三视角 │ 侧面   │ 背面   │
│        │        │        │        │
└────────┴────────┴────────┴────────┘

bash

undefined

Generate front view

生成正面视图

infsh app run falai/flux-dev-lora --input '{ "prompt": "character design, front view, young woman with short asymmetric red hair, bright green eyes, wearing navy blue bomber jacket over white graphic tee, dark jeans, red sneakers, standing in neutral pose, full body, clean white background, concept art, sharp details", "width": 768, "height": 1024 }' --no-wait

Generate 3/4 view (same description)

生成四分之三视角（相同描述）

infsh app run falai/flux-dev-lora --input '{ "prompt": "character design, three-quarter view, young woman with short asymmetric red hair, bright green eyes, wearing navy blue bomber jacket over white graphic tee, dark jeans, red sneakers, standing, full body, clean white background, concept art, sharp details", "width": 768, "height": 1024 }' --no-wait

Generate side view

生成侧面视图

infsh app run falai/flux-dev-lora --input '{ "prompt": "character design, side profile view, young woman with short asymmetric red hair, bright green eyes, wearing navy blue bomber jacket over white graphic tee, dark jeans, red sneakers, standing, full body, clean white background, concept art, sharp details", "width": 768, "height": 1024 }' --no-wait

Generate back view

生成背面视图

infsh app run falai/flux-dev-lora --input '{ "prompt": "character design, back view, young woman with short asymmetric red hair, wearing navy blue bomber jacket over white graphic tee, dark jeans, red sneakers, standing, full body, clean white background, concept art, sharp details", "width": 768, "height": 1024 }' --no-wait

Stitch into reference sheet

拼接成参考表

infsh app run infsh/stitch-images --input '{ "images": ["front.png", "three-quarter.png", "side.png", "back.png"], "direction": "horizontal" }'

undefined

infsh app run infsh/stitch-images --input '{ "images": ["front.png", "three-quarter.png", "side.png", "back.png"], "direction": "horizontal" }'

undefined

2. Expression Sheet

2. 表情表

Shows the character's face with different emotions:

┌────────┬────────┬────────┐
│NEUTRAL │ HAPPY  │ ANGRY  │
│        │        │        │
├────────┼────────┼────────┤
│  SAD   │SURPRISE│THINKING│
│        │        │        │
└────────┴────────┴────────┘

Minimum 6 expressions: neutral, happy, angry, sad, surprised, thinking.

bash

undefined

展示角色的不同面部情绪：

┌────────┬────────┬────────┐
│ 中性   │ 开心   │ 愤怒   │
│        │        │        │
├────────┼────────┼────────┤
│ 悲伤   │ 惊讶   │ 思考   │
│        │        │        │
└────────┴────────┴────────┘

至少包含6种表情：中性、开心、愤怒、悲伤、惊讶、思考。

bash

undefined

Neutral

中性表情

infsh app run falai/flux-dev-lora --input '{ "prompt": "character portrait, close-up face, young woman with short red hair and green eyes, neutral calm expression, clean white background, concept art, consistent character design", "width": 512, "height": 512 }' --no-wait

Happy

开心表情

infsh app run falai/flux-dev-lora --input '{ "prompt": "character portrait, close-up face, young woman with short red hair and green eyes, warm genuine smile, happy expression, clean white background, concept art, consistent character design", "width": 512, "height": 512 }' --no-wait

Angry

愤怒表情

infsh app run falai/flux-dev-lora --input '{ "prompt": "character portrait, close-up face, young woman with short red hair and green eyes, furrowed brows, angry determined expression, clean white background, concept art, consistent character design", "width": 512, "height": 512 }' --no-wait

(Continue for sad, surprised, thinking...)

（继续生成悲伤、惊讶、思考表情...）

undefined

undefined

3. Outfit/Costume Sheet

3. 服装表

Multiple outfits for the same character:

Outfit	Description
Casual	Bomber jacket, t-shirt, jeans
Work	Blazer, button-down, slacks
Athletic	Sports bra, leggings, running shoes
Formal	Evening dress, heels

同一角色的多套服装：

服装类型	描述
休闲装	飞行员夹克、T恤、牛仔裤
工作装	西装外套、衬衫、西裤
运动装	运动内衣、紧身裤、跑鞋
正装	晚礼服、高跟鞋

4. Color Palette Sheet

4. 调色板表

Document exact colors for consistency:

CHARACTER: Maya Chen

Skin:    ████ #F5D0A9 (warm beige)
Hair:    ████ #C0392B (auburn red)
Eyes:    ████ #27AE60 (emerald green)
Jacket:  ████ #2C3E50 (navy blue)
T-shirt: ████ #ECF0F1 (off-white)
Jeans:   ████ #34495E (dark slate)
Shoes:   ████ #E74C3C (bright red)

记录精确颜色以保证一致性：

角色：Maya Chen

肤色:    ████ #F5D0A9（暖米色）
发色:    ████ #C0392B（赤褐色）
瞳色:    ████ #27AE60（祖母绿）
夹克:    ████ #2C3E50（藏青色）
T恤:     ████ #ECF0F1（米白色）
牛仔裤:  ████ #34495E（深灰蓝）
鞋子:    ████ #E74C3C（亮红色）

The Description Anchor Technique

描述锚点技巧

The most practical consistency technique: write a 50+ word detailed description and reuse it exactly in every prompt.

最实用的一致性技巧：撰写一段50字以上的详细描述，并在所有提示词中完全复用。

Template

模板

[age] [gender] with [hair: color, length, style], [eye color] eyes,
[skin tone], [facial features: any distinctive marks],
wearing [top: specific color and style], [bottom: specific color and style],
[shoes: specific color and style], [accessories: specific items]

[年龄] [性别]，拥有[头发：颜色、长度、样式]，[瞳色]眼睛，
[肤色]，[面部特征：独特标记]，
穿着[上衣：具体颜色和样式]，[下装：具体颜色和样式]，
[鞋子：具体颜色和样式]，[配饰：具体物品]

Example

示例

young woman in her mid-twenties with short asymmetric auburn red hair
swept to the right side, bright emerald green eyes, light warm skin
with a small beauty mark below her left eye, wearing a fitted navy
blue bomber jacket with silver zipper over a white crew-neck t-shirt,
dark slate slim jeans, and bright red canvas sneakers, small silver
stud earrings

Use this exact block in EVERY prompt for this character, only changing the action/pose/scene.

二十五岁左右的年轻女性，留着偏分至右侧的短款不对称赤褐色头发，
明亮的祖母绿眼睛，浅暖肤色，左眼下方有一颗小美人痣，
穿着带有银色拉链的修身藏青色飞行员夹克，内搭白色圆领T恤，
深灰蓝修身牛仔裤，脚穿亮红色帆布运动鞋，佩戴小巧的银色耳钉

在该角色的每一条提示词中都使用这段完全相同的描述，仅修改动作/姿势/场景。

Proportion Guide

比例指南

Style	Head-to-Body Ratio	Best For
Realistic	7.5 : 1	Film, photorealistic
Heroic	8 : 1	Superheroes, action
Anime/Manga	5-6 : 1	Japanese animation style
Stylized	4-5 : 1	Western animation
Chibi/Super-deformed	2-3 : 1	Cute, comedic, mascots

Include proportion style in your prompts: "realistic proportions" vs "anime style proportions" vs "chibi proportions"

风格	头身比	适用场景
写实风格	7.5 : 1	影视、照片级写实
英雄风格	8 : 1	超级英雄、动作题材
动漫/漫画风格	5-6 : 1	日式动画风格
风格化	4-5 : 1	西式动画
Q版/超变形	2-3 : 1	可爱风、喜剧、吉祥物

在提示词中注明比例风格：如“写实比例”、“动漫风格比例”、“Q版比例”

Using LoRA for Consistency

使用LoRA实现一致性

For projects requiring many images of the same character, train a LoRA:

bash

undefined

对于需要大量该角色图像的项目，训练一个LoRA：

bash

undefined

Use FLUX with a character LoRA

结合角色LoRA使用FLUX

infsh app run falai/flux-dev-lora --input '{ "prompt": "maya_chen character, sitting at a cafe reading a book, warm afternoon light, candid photography style", "loras": [{"path": "path/to/maya-chen-lora.safetensors", "scale": 0.8}] }'


**LoRA Training Tips:**
- Need 10-20 reference images of the character (consistent style)
- Train on specific trigger word (e.g., "maya_chen")
- Scale 0.7-0.9 balances consistency with prompt flexibility
- Lower scale = more creative freedom, higher = more strict matching

infsh app run falai/flux-dev-lora --input '{ "prompt": "maya_chen character, sitting at a cafe reading a book, warm afternoon light, candid photography style", "loras": [{"path": "path/to/maya-chen-lora.safetensors", "scale": 0.8}] }'


**LoRA训练技巧：**
- 需要10-20张该角色的参考图像（风格一致）
- 训练时设置专属触发词（如“maya_chen”）
- 0.7-0.9的权重可平衡一致性与提示词灵活性
- 权重越低 = 创作自由度越高，权重越高 = 与参考匹配度越严格

Common Consistency Failures

常见一致性问题

Issue	Why It Happens	Mitigation
Hair color drift	Model interprets "red hair" differently each time	Use specific shade: "auburn red #C0392B"
Eye color change	Low priority in generation	Mention eye color early in prompt
Outfit inconsistency	Model fills in details creatively	Describe every clothing item explicitly
Age shift	Vague age description	Use "mid-twenties" not "young"
Face structure change	Different generations = different faces	Use LoRA or same seed base
Proportion shift	Style interpretation varies	Specify "7.5 head proportions"

问题	原因	解决方法
发色偏差	模型对“红发”的理解每次不同	使用具体色号：“赤褐色 #C0392B”
瞳色变化	生成优先级较低	在提示词开头提及瞳色
服装不一致	模型会自主补充细节	明确描述每一件衣物
年龄偏差	年龄描述模糊	使用“二十五岁左右”而非“年轻”
面部结构变化	每次生成的面部不同	使用LoRA或相同种子的基础图像
比例偏差	风格解读不一致	明确标注“7.5头身比例”

Character Bible Template

角色设定集模板

For ongoing projects, maintain a character bible document:

markdown

undefined

对于长期项目，维护一份角色设定集文档：

markdown

undefined

Character: Maya Chen

角色：Maya Chen

Visual Description (use in all prompts)

视觉描述（所有提示词中使用）

young woman in her mid-twenties with short asymmetric auburn red hair... [full 50+ word anchor description]

二十五岁左右的年轻女性，留着偏分至右侧的短款不对称赤褐色头发... [完整的50字以上锚点描述]

Color Palette

调色板

Skin: #F5D0A9
Hair: #C0392B
Eyes: #27AE60
Primary outfit: Navy #2C3E50
Accent: Red #E74C3C

肤色：#F5D0A9
发色：#C0392B
瞳色：#27AE60
主服装：藏青色 #2C3E50
强调色：亮红色 #E74C3C

Personality Notes (for expression/pose choices)

性格说明（用于表情/姿势选择）

Confident but approachable
Default expression: slight curious smile
Gestures: talks with hands, leans forward when interested

自信且平易近人
默认表情：略带好奇的微笑
手势：说话时会用手辅助，感兴趣时会前倾身体

Style Keywords

风格关键词

concept art, clean lines, sharp details, [art style reference]

概念美术、简洁线条、细节清晰、[美术风格参考]

LoRA (if trained)

LoRA（若已训练）

Path: ./loras/maya-chen-v2.safetensors Trigger: maya_chen Recommended scale: 0.8

undefined

路径：./loras/maya-chen-v2.safetensors 触发词：maya_chen 推荐权重：0.8

undefined

Common Mistakes

常见误区

Mistake	Problem	Fix
Vague descriptions	Different character every time	50+ word detailed anchor
Inconsistent prompt structure	Varying emphasis = varying results	Same structure, only change action/scene
Generating one view only	Can't use character in different contexts	Create full turnaround reference
No color documentation	Colors drift across generations	Record exact hex codes
Skipping expression sheet	Character feels one-dimensional	Generate 6+ expressions
Not using LoRA for big projects	Inconsistency compounds	Train LoRA for 10+ image projects

误区	问题	解决方法
描述模糊	每次生成的角色都不同	撰写50字以上的详细锚点描述
提示词结构不一致	重点变化导致结果变化	使用相同结构，仅修改动作/场景
仅生成单一视角	无法在不同场景中使用角色	创建完整的多角度视图参考表
未记录颜色	颜色在多次生成中偏差	记录精确的十六进制色码
未制作表情表	角色显得单薄	生成6种以上表情
大型项目未使用LoRA	不一致问题会累积	对于需要10张以上图像的项目，训练LoRA

character-design-sheet

Original

Translation

Character Design Sheet

角色设计参考表

Quick Start

快速开始

Generate a character concept

生成角色概念

The Consistency Problem

一致性问题

Solutions (Ranked by Effectiveness)

解决方案（按有效性排序）

Reference Sheet Types

参考表类型

1. Turnaround Sheet

1. 多角度视图表

Generate front view

生成正面视图

Generate 3/4 view (same description)

生成四分之三视角（相同描述）

Generate side view

生成侧面视图

Generate back view

生成背面视图

Stitch into reference sheet

拼接成参考表

2. Expression Sheet

2. 表情表

Neutral

中性表情

Happy

开心表情

Angry

愤怒表情

(Continue for sad, surprised, thinking...)

（继续生成悲伤、惊讶、思考表情...）

3. Outfit/Costume Sheet

3. 服装表

4. Color Palette Sheet

4. 调色板表

The Description Anchor Technique

描述锚点技巧

Template

模板

Example

示例

Proportion Guide

比例指南

Using LoRA for Consistency

使用LoRA实现一致性

Use FLUX with a character LoRA

结合角色LoRA使用FLUX

Common Consistency Failures

常见一致性问题

Character Bible Template

角色设定集模板

Character: Maya Chen

角色：Maya Chen

Visual Description (use in all prompts)

视觉描述（所有提示词中使用）

Color Palette

调色板

Personality Notes (for expression/pose choices)

性格说明（用于表情/姿势选择）

Style Keywords

风格关键词

LoRA (if trained)

LoRA（若已训练）

Common Mistakes

常见误区

Related Skills

相关技能