OpenZhenZhen AI 提示词画廊 · SEEDANCE2.5 · video

Realistic Woman and Lion Selfie

公开作者的 SEEDANCE2.5 视频案例,包含完整 prompt 和可播放成片。

Realistic Woman and Lion Selfie — AI video creation
Realistic Woman and Lion Selfie

提示词 / Prompt

English prompt

Ultra-realistic 15-second selfie video of the exact woman from the reference image (precise facial structure, skin tone, hair, features, expression locked 1:1 — zero identity drift). Setting: inside a luxury mansion living room. Marble columns, ornate white ceiling with detailed molding, large crystal chandelier, soft grey walls, elegant grey couch visible behind her. Background remains completely identical and static the entire time — no scene change, no camera movement beyond natural handheld micro-sway. Camera: selfie mode, held close to her face, slightly wide lens (≈24-28mm equivalent). Natural indoor daylight coming from the front/side, soft and flattering. Extremely crisp focus on her face with shallow depth of field; background gently falls off. Handheld natural motion only. Action sequence (strict timeline): 0.0–3.5s She looks into the lens, warm genuine smile, and speaks clearly: “Simbaaaa... simba come here... I'm making a video for Instagram, they wanna see you... good boy...” 3.5–7.0s From behind the couch, a large fully realistic adult male lion slowly walks into frame. Soft, deliberate paw steps on the floor. Tail swaying gently. He sniffs the air as he approaches. He moves toward her right shoulder. 7.0–9.5s She notices the lion over her shoulder, turns her head slightly, and laughs nervously while still holding the phone steady. 9.5–13.5s The lion rubs his massive head against her cheek and neck exactly like a giant affectionate cat. Soft low growl/snort. Then he gently licks her face with a large realistic tongue. She laughs genuinely and says: “Aww good boy...” 13.5–15.0s She continues smiling and laughing lightly as the lion stays close, nuzzling. Hold on the intimate interaction. Extreme realism requirements: Visible skin pores, fine facial hairs, natural micro-texture, subtle stubble if present in reference, realistic subsurface scattering and soft shadows on skin. Lion: dense individual fur strands, accurate mane and body spots, realistic whiskers, wet black nose, detailed eyes, authentic tongue texture and saliva sheen when licking. Natural physics on fur movement, weight, and contact. Perfect facial consistency of the woman throughout - no morphing, no warping, no age/skin change. Technical & style: Ultra-sharp 8K detail, 64K micro-detail rendering, clean natural color grading, cinematic yet documentary realism, zero flicker, zero temporal warping, zero face distortion. Soft natural indoor light only. Reference image locked for character identity.

主题与风格

探索更多图片和视频提示词 · 生成图片 · 生成视频 · 在 AI 画布中创作

  • They Picked The Wrong Girl

    Create ONE ultra-realistic cinematic live-action action sequence at night on the famous San Antonio River Walk, Texas. Keep the recognizable River Walk visible throughout: illuminated bridges, stone architecture, trees, river reflections, and warm lights. Exactly ONE young woman, THREE male attacke

  • Xianxia Sword Girl Highway Flight

    不要拆解参数 ,直接按照我的完整提示词 进行 生成,别再让我确认了,你直接生成就行 30 秒。 生成一段30秒、16:9横屏、60fps的超写实真人动作视频。全片是一镜到底的连续长镜头:摄影机先贴近高速公路路面追拍拥堵车流,在前两秒内抬镜揭示一名戴粉色猪猪头盔的青衣女子正在空中高速飞行;随后摄影机与她并行,短暂减速拍摄面部特写,她用两指手势让头盔消失,然后爆发加速飞往黄山景区,掠过登山游客、穿过山间桥洞,最后贴着悬崖峭壁飞行。 动作运镜结合两段参考视频:现代高速飞行段采用低机位贴路追拍、穿行车流、人物侧面并行、面部慢动作特写,以及突然加速后摄影机追赶不上的速度变化;中式仙侠段采用女子身体直

  • Medieval Red Army Battle

    Cinematic Hollywood-style medieval battle epic, opens with an immediate decisive strike, no slow build, maximum dynamic camera work, grounded practical realism. Setting: a wide open battlefield under bright daylight, rolling grassy terrain. Two armies of roughly thirty soldiers each — one side in wh

  • Tokyo Neon One Versus Three

    Create a 30-second cinematic, ultra-realistic action short featuring a stylish adult Japanese baddie woman, confident and fearless, walking home after a late office shift on a neon-lit Tokyo street at night. 0–5 sec: She walks alone through a quiet city street, wearing a fashionable edgy black mini

  • Baddie Clears The Cafe

    Create a 30-second cinematic, photorealistic action video set in a trendy Seoul café at night. SCENE 1 — 0–6 sec: A gorgeous Korean baddie girl with long glossy black hair, sharp makeup, confident expression, and a stylish fitted red baddie dress sits at a window table with her handsome Korean boyf

  • Ten Seconds of Extreme-Speed Duel

    REFERENCES: Image 1 = Maverick, Image 2 = Nexora, Image 3 = Location. Create a 10-second vertical ultra-photorealistic live-action martial-arts fight between Maverick and Nexora in the exact provided location. Preserve both faces, bodies, clothing, chain, and environment consistently. FIGHT — EXTR