2x2 网格,16:9,为 4 位著名物理学家执行以下操作:class VisualPunLogo:
分享链接 · https://openimages.ajiang.me/p/2095440664133849274/

color_palette · high-contrastcolor_palette · white-minimalcomposition · multi-panelmood · playfulquality_flags · embedded-textquality_flags · multi-panel-layout
提示词(原文,逐字保留)
2x2 grid, 16:9, do this for 4 famous physicists : class VisualPunLogo:
def __init__(self, SUBJECT, WORD):
s = infer(SUBJECT, WORD) # boundary, target_letter, prop, palette
self.s = s
def gestalt_trace(self): # V₁: Text = Object boundary
return curve_text_to_boundary(self.s.word, self.s.boundary)
def semantic_morph(self, text_mesh): # V₂: Letter = Prop (exactly one)
target_idx = find_letter(text_mesh, self. s. target)
return morph_stroke(text_mesh, target_idx, into=self.s.prop)
def render(self):
base = self.gestalt_trace()
final = self.semantic_morph(base)
return vector_logo(final,
palette=self.s.palette, # V₄: 2-4 spot colors
bg=pure_white,
style=minimalist_3D_flat)
def verify(self, img):
assert reads_as_word(img, self.s.word) and reads_as_object(img, self.s.subject), "V₃: dual-read fails"
assert morph_count(img) == 1, "V₂: too many morphs"
assert prop_belongs_to_subject(img, self.s.prop, self.s.subject), "V₅: category error"
assert is_flat_vector(img) and len(colors(img)) <= 4 and bg_is_white(img), "V₄: vector discipline"
VisualPunLogo(SUBJECT, WORD).render()
中文译文
2x2 网格,16:9,为 4 位著名物理学家执行以下操作:class VisualPunLogo:
def __init__(self, SUBJECT, WORD):
s = infer(SUBJECT, WORD) # boundary, target_letter, prop, palette
self.s = s
def gestalt_trace(self): # V₁: Text = Object boundary
return curve_text_to_boundary(self.s.word, self.s.boundary)
def semantic_morph(self, text_mesh): # V₂: Letter = Prop (exactly one)
target_idx = find_letter(text_mesh, self. s. target)
return morph_stroke(text_mesh, target_idx, into=self.s.prop)
def render(self):
base = self.gestalt_trace()
final = self.semantic_morph(base)
return vector_logo(final,
palette=self.s.palette, # V₄: 2-4 spot colors
bg=pure_white,
style=minimalist_3D_flat)
def verify(self, img):
assert reads_as_word(img, self.s.word) and reads_as_object(img, self.s.subject), "V₃: dual-read fails"
assert morph_count(img) == 1, "V₂: too many morphs"
assert prop_belongs_to_subject(img, self.s.prop, self.s.subject), "V₅: category error"
assert is_flat_vector(img) and len(colors(img)) <= 4 and bg_is_white(img), "V₄: vector discipline"
VisualPunLogo(SUBJECT, WORD).render()
来源与署名
原文由 @Gdgtify 发布在 X:查看原推文。 本页逐字保留原文并提供机器翻译的中文解读;版权归原作者所有。
同工具的更多提示词
- 严格4:5竖版构图比例。创作一张高级感、超写实的微缩旅行明信片,灵感来自[LOCATION],打造电影感“城市在手心”的画面。一只写实的人类手掌轻柔地托起一块…@Goodmanprotocol
- 创建一个照片写实主义的电影感街景,画面中一部巨大的现代智能手机斜放在城市人行道上。一位时尚女性自然地坐在这部超大的手机上,身穿修身棕色长袖上衣、高腰米色长裤和…@Zara__Ai
- 为我生成一张图片:一位女性身穿棕色抹胸上衣、豹纹层叠迷你裙、黑色金属腰带、棕色皮革项圈,留着长长的棕色波浪卷发。她妆容精致,风格偏向Y2K(千禧辣妹风),脚踩…@kriaszone
- 9:16竖幅,真实手机摄影。韩国成年女性,白皙自然肤质,深色头发随意扎成低松发髻,几缕湿发垂落脸侧;米白色细肩带针织上衣,宽松米白针织开衫自然滑落肩部,极细金…@DeepBlueZer0
- 真实的品牌名称,复制的源文本,长标题文案,杂乱的布局,弱化的产品焦点,廉价糖果广告风格,低细节包装纸,人工水果纹理,暗淡的粉色调,凌乱的飞溅,塑料感十足的意式…@ou_zhen599
- [GPT Image 2 Prompt] | ar 9:16 4K High Created on @TopviewAIhq @topviewaijp #To…@woleswoosh