12位日本美女写实发型展示网格

A photo-realistic hairstyle collection presented by 12 different Japanese beauties aged 25-35. Change the facial contours, eye/nose/mouth shapes, hair textures, and atmospheres of all 12 people separately, naturally expressing them as distinctly different female models. Hairstyle length, cut, color, bangs, parting, hair flow, curl, and styling are completely random every time. All 12 types are different hairstyles; it is forbidden to repeat or have overly similar hairstyles just by changing hair color. Specific hairstyles are not specified and are freely conceived by the AI. The shooting range is a portrait centered on the face. Shoot a range that allows the entire face and hairstyle characteristics to be seen clearly, ensuring the hair is not unnaturally cropped. Face orientation, expression, gaze, and background are also randomly changed in each of the 12 photos. However, avoid extreme angles where the hairstyle is hard to identify or poses that cover the face. Backgrounds are hair salons, interiors, streets, windowsides with natural light, etc.; the AI freely chooses locations suitable for the hairstyles. All 12 photos use different backgrounds. Realistic portrait photos taken with natural light or actual ambient light. Accurately represent the texture, volume, bundles, natural luster, and the dynamics of bangs and hair ends of each strand of hair. Skin texture, facial bone structure, the relationship with ears and neck, and the hairline have no unnatural parts. Light sources, color temperature, and shadows of the characters and backgrounds are consistent, avoiding a sense of incongruity like composite photos. Do not use excessive skin beautification or CG-style lighting. The overall image is a vertical 9:16 aspect ratio, a uniform grid of 3 columns x 4 rows. Display white text with black border numbers 1-12 in the upper left of each image. Combine all 12 into one image without adding redundant text or titles. With each generation, the combination of models, hairstyles, expressions, face orientation, and backgrounds will change.

由25〜35岁的不同日本美女12人呈现的、照片级真实发型收藏。

将12人全体的面容轮廓、眼鼻口形状、发质、氛围分别改变，作为明显不同的女性模特自然表现。

发型长度、剪裁、颜色、前刘海、分线、毛流向、卷曲、造型每次完全随机。12种全部为不同发型，禁止仅改变发色重复或过于相似的发型。具体发型不指定，由AI自由构思。

拍摄范围以脸部为中心的肖像。拍摄能看清整个脸部和发型特征的范围，确保头发不被不自然地裁切。

脸部朝向、表情、视线、背景也在12张照片中分别随机改变。但要避免发型难以辨认的极端角度或遮挡脸部的姿势。

背景为美发沙龙、室内、街头、自然光照射的窗边等，AI自由选择适合发型的场所。12张照片全部使用不同背景。

以自然光或实际环境光拍摄的真实肖像照片。精确表现每根头发的质感、发量、发束、自然光泽、前刘海及发梢的动态。

皮肤质感、脸部骨骼、与耳朵及脖子的位置关系、发际线无不自然之处。人物与背景的光源、色温、阴影保持一致，避免合成照片般的违和感。不使用过度美肌处理或CG风格的照明。

图像整体为纵向9:16，3列×4行的均匀网格。

每张图像左上方显示1〜12的白色文字、黑色边框编号。将12张全部合成一张图像，不添加多余文字或标题。

每次生成时，模特、发型、表情、脸部朝向、背景的组合都会发生变化。