使用原图创建一个虚拟的Vtuber及其直播画面。 Vtuber的发型和服装将忠实还原原图。 Vtuber的画面为2.5D画质，无需完美还原原图风格，但需要适度的立体感。 Vtuber的表情和姿势可以与原图有所差异。请让Vtuber手持游戏手柄。 仅将Vtuber的上半身放置在屏幕右下方。将正在游玩的游戏直播画面放置在屏幕中央。将聊天画面放置在屏幕左侧。 游戏画面的屏幕比例设置为较大，Vtuber的上半身显示较小。 完全忽略原图的背景和姿势。 在屏幕的上下添加虚拟的直播平台和浏览器UI。 生成图像的纵横比与原始图像的纵横比无关。

{ "generation_request": { "meta_data": { "tool": "NanoBanana Pro", "task_type": "text_to_image_ultra_photoreal_beauty_contact_sheet_product_campaign", "version": "v1.1", "priority": "high" }, "output_settings": { "aspect_ratio": "4:5", "orientation": "portrait", "resolution": "ultra_high_res", "render_style": "ultra_photoreal_beauty_editorial", "sharpness": "crisp_but_natural", "film_grain": "none_or_minimal", "color_grade": "clean_true_to_life" }, "creative_prompt": { "scene_summary": { "en": "Ultra-photoreal beauty campaign contact sheet on a pure white seamless background. A 3x3 grid collage of nine panels featuring ONE model with hair worn down: voluminous, glossy, long dark waves with soft movement and lift at the roots. The model has a lean, fit, runway-like physique and an attractive high-fashion look. Skin is luminous with realistic texture and refined editorial retouch (no plastic smoothing). Makeup: defined brows, clean eyeliner, soft bronzy blush, and bold red lipstick.\n\nProduct must stay consistent: a sleek chrome lip product tube and doe-foot applicator with the brand name '{argument name="brand name" default="BAGEL"}' printed clearly on the tube in a bold modern sans-serif. The tube shape, cap, chrome finish, and label placement remain consistent across all panels. The word '{argument name="brand name" default="BAGEL"}' is readable, spelled correctly, centered, and aligned on the tube in every panel where the product is visible. No other text or brands anywhere.\n\nPanel variety (9 frames):\n1) Model holds the chrome tube horizontally over the eyes like a visor; lips puckered.\n2) Close-up applying the red lipstick with the doe-foot applicator; tube visible with '{argument name="brand name" default="BAGEL"}'.\n3) Model laughing wide-mouth; tube held near face.\n4) Beauty angle portrait holding tube near chin; confident gaze.\n5) Extreme close-up smile with teeth; tube held lightly between teeth; '{argument name="brand name" default="BAGEL"}' readable.\n6) Wink + playful tongue-out; holding applicator and tube; '{argument name="brand name" default="BAGEL"}' readable.\n7) Applying lipstick again with elegant hand pose; manicure clean nude/white.\n8) Kissy face wink; cheek highlight strong and glossy.\n9) Tight crop smile; crisp lip texture and clean edges.\n\nLighting: high-end beauty studio lighting, bright and even, soft specular highlights on chrome packaging, realistic shadows, face and product both tack sharp. Composition: clean margins, consistent spacing between panels, magazine-quality production. No watermark, no extra logos." }, "composition": { "layout": "3x3 grid collage, consistent white gutters, clean alignment", "framing": "mix of tight close-ups and medium close-ups; product always readable when present", "focus": "face sharp; BAGEL text sharp; background pure white" }, "lighting_and_camera": { "lighting": "beauty studio: large soft key + fill + subtle rim; bright, clean, shadow-controlled", "lens": "85

Design a viral video thumbnail using the person from Image 1. Face Consistency : Keep the person's facial features exactly the same as Image 1 , but change their expression to look excited and surprised . Action : Pose the person on the left side, pointing their finger towards the right side of the frame. Subject : On the right side, place a high-quality image of [a delicious avocado toast]. Graphics : Add a bold yellow arrow connecting the person's finger to the toast. Text : Overlay massive, pop-style text in the middle: '3分钟搞定!' (Done in 3 mins!). Use a thick white outline and drop shadow. Background : A blurred, bright kitchen background. High saturation and contrast.

使该对象透明

在整个图片上反复覆盖“TRUMP”这个词。

主体：虚拟全息角色 [CHARACTER]，悬浮于现代桌面上直径 120 毫米的圆形全息投影仪上方。 投影源规则： - 如果输入参考对象是 3D 物体，则在投影仪旁边放置一台桌面 3D 扫描仪。 将参考对象放置在扫描板上。 投影仪上方的全息图由扫描对象生成。 - 如果输入参考对象是 2D 图像，则在桌面上放置一台带显示器的现代 PC。 在显示器屏幕上显示参考图像。 投影仪上方的全息图由该屏幕内容生成。 全息图渲染规则： - 角色始终显示为半透明的体积图像，背景隐约可见。 - 无光束、无粒子、无实体雕像表面。 - 均衡的解剖结构（头身比例为 1/7-1/8），比例正确。 - 姿势自然，轮廓清晰。 - 头发、衣服褶皱和配饰可见但半透明。 - 面部清晰富有表现力，在 1000 像素裁剪下清晰可读。 - 无版权角色、无品牌设计、无 IP 标识。 环境：现代办公桌，带投影仪底座 + 辅助设备（扫描仪或显示器）。 相机：85-100 毫米镜头，3/4 英雄角，眼平，f/11-f/16，ISO100，三脚架。 照明：桌面柔和照明；全息人物仅由体积光定义。 背景：无缝黑色工作室，反射微妙。 输出：4:5，2048×2560。 负片：无文字、无水印、无标识、无品牌、版权角色、系列 IP、商标设计、树脂、PVC、实体雕像、不透明表面、玩具光泽、光束、扫描线、点、失真、多余数字。 采样：确定性，Seed=12345，Temperature=0。

把我的X账户做成90年代软盘