<instruction> Do this for [Animal / Creature / Pet] practicing [Human Skill / Martial Art / Discipline]:
1. Skill Transfer:
Input A is an animal or creature.
Input B is a human discipline, martial art, craft, sport, or skill.
Analyze both inputs and infer:
Species Motion Traits: the natural posture, balance, limb movement, paw/wing/claw use, and body proportions of Input A.
Discipline Mechanics: the essential movement or technique associated with Input B (stance, strike, block, craft motion, or tool use).
Training Interaction: determine how a human trainer, instructor, or environment would realistically guide or test the skill.
Scale & Equipment: infer the correct training props, uniforms, or equipment sized appropriately for the animal.
2. Container
Goal: a believable practice or training scene.
The environment should be the natural location where the discipline would occur:
dojo, gym, workshop, field, training hall, studio, kitchen station, laboratory bench, etc.
The environment should contain recognizable training objects or equipment relevant to Input B.
3. Scene
Input A actively performs the discipline in a recognizable stance or action.
A trainer, instructor, or observer may assist by holding equipment, guiding posture, or presenting a challenge (such as a target or object).
The moment should capture the key action of the skill: striking, crafting, balancing, lifting, cutting, shaping, blocking, or demonstrating technique.
4. Realism Constraint:
preserve realistic anatomy of Input A (fur patterns, whiskers, paws, tail movement, posture).
any clothing or gear must support the activity rather than dominate the scene (training belt, small uniform, gloves, apron, etc.).
tools and objects must be physically scaled so the action feels believable.
5. Visual Syntax
Camera: candid mid-range shot like a paused video frame or training moment
Lighting: natural indoor lighting or soft gym lighting
Detail: realistic fur, equipment textures, and environmental details
Mood: playful but believable, skill demonstration moment
6. Flexibility Constraint:
This should work for any combination such as:
cat practicing karate
dog lifting weights
raccoon woodworking
bird painting
rabbit gardening
The system must infer correct props, stance, and environment from the variable inputs without hardcoding specific animals or disciplines.
Output: ONE image, vertical or square aspect ratio, realistic candid training moment of an animal performing a human discipline with believable environment and props.
</instruction>
<instruction> 针对正在练习 [人类技能 / 武术 / 学科] 的 [动物 / 生物 / 宠物] 执行以下操作:
1. 技能转移:
输入 A 是动物或生物。
输入 B 是人类学科、武术、手工艺、运动或技能。
分析两个输入并推断:
物种运动特征:输入 A 的自然姿势、平衡、肢体运动、爪/翅膀/利爪的使用以及身体比例。
学科机制:与输入 B 相关的基本动作或技术(姿态、打击、格挡、制作动作或工具使用)。
训练互动:确定人类训练师、教练或环境将如何真实地指导或测试技能。
尺寸与设备:推断适合该动物尺寸的正确训练道具、制服或设备。
2. 容器
目标:一个可信的练习或训练场景。
环境应是该学科发生的自然场所:
道场、健身房、工作室、田野、训练馆、画室、厨房工作台、实验室工作台等。
环境应包含与输入 B 相关的可识别训练物品或设备。
3. 场景
输入 A 以可识别的姿态或动作积极地执行该学科。
训练师、教练或观察者可以通过手持设备、指导姿势或提出挑战(例如目标或物体)来协助。
该时刻应捕捉技能的关键动作:打击、制作、平衡、举起、切割、塑形、格挡或演示技术。
4. 真实性约束:
保留输入 A 的真实解剖结构(毛发图案、胡须、爪子、尾巴摆动、姿势)。
任何衣物或装备都必须支持活动,而不是主导场景(训练腰带、小制服、手套、围裙等)。
工具和物体必须按实际比例缩放,以便动作感觉真实可信。
5. 视觉语法
相机:像暂停的视频帧或训练时刻一样的抓拍中景镜头
光线:自然室内光线或柔和的健身房光线
细节:逼真的毛发、设备纹理和环境细节
情绪:活泼但可信,技能展示时刻
6. 灵活性约束:
这应该适用于任何组合,例如:
练习空手道的猫
举重狗
浣熊木工
鸟类绘画
兔子园艺
系统必须根据可变输入推断出正确的道具、姿态和环境,而无需硬编码特定的动物或学科。
输出:一张图片,垂直或方形长宽比,逼真抓拍的动物在可信环境和道具下进行人类学科训练的瞬间。
</instruction>
Comments & Questions (0)
No comments yet
Be the first to comment!