ساخت کلیپ تولد با هوش مصنوعی: آموزش عملی همراه ۶ پرامپت رایگان

ساخت کلیپ تولد با هوش مصنوعی فقط به گذاشتن چند عکس کنار هم و اضافه کردن موسیقی محدود نمی‌شود. ابزارهای جدید می‌توانند یک عکس ساده را به ویدیویی کوتاه و متحرک تبدیل کنند؛ مثلا فرد داخل تصویر لبخند بزند، به کیک نگاه کند، شمع‌ها حرکت کنند یا دوربین به‌آرامی به سوژه نزدیک شود. حتی می‌توان جلوه‌هایی مثل ریزش کاغذهای رنگی و تزئینات تولد را هم به صحنه اضافه کرد.

در ادامه این محتوا به شما می‌گوییم:

• چطور عکس مناسب برای ساخت کلیپ انتخاب کنید.
• چگونه با نوشتن یک پرامپت دقیق، حرکت موردنظرتان را برای هوش مصنوعی توضیح دهید.
• ۶ پرامپت ساخت کلیپ تولد ترند را با قابلیت شخصی‌سازی در اختیارتان قرار می‌دهیم.
• و ۶ ابزار برتر را برای ساخت کلیپ تولد با هوش مصنوعی به شما معرفی می‌کنیم.

قبل از ساخت کلیپ تولد با هوش مصنوعی چه چیزهایی آماده کنیم؟

کیفیت کلیپ تا حد زیادی به عکسی بستگی دارد که در اختیار هوش مصنوعی قرار می‌دهید. اگر تصویر اولیه تار باشد، چهره جزئیات کافی نداشته باشد یا دست‌ها در وضعیت غیر طبیعی قرار گرفته باشند، احتمال بروز این مشکلات در ویدیوی نهایی هم وجود دارد.

قبل از شروع، بهتر است موارد زیر را آماده کنید:

• عکس مناسب: تصویری واضح با جزئیات کافی از چهره و سوژه اصلی انتخاب کنید.
• سبک ویدیو: مشخص کنید کلیپ را با چه فضایی می‌خواهید؛ مثلا دوستانه، خانوادگی، فانتزی یا لاکچری.
• ابعاد ویدیو: محل انتشار کلیپ را مشخص کنید تا نسبت تصویر مناسب را انتخاب کنید. به‌‌عنوان مثال اگر قصد دارید کلیپ را در استوری یا ریلز اینستاگرام منتشر کنید، بهتر است از ابتدا خروجی عمودی با نسبت تصویر 9:16 را در نظر بگیرید.
• ایده حرکت: مشخص کنید دوست دارید چه اتفاقی در ویدیو بیفتد؛ مثلا لبخند زدن، فوت کردن شمع‌ها یا حرکت دوربین.

پرامپت عکس تولد با هوش مصنوعی؛ ایده‌های جذاب و حرفه‌ای برای عکس تولد

آموزش ساخت کلیپ تولد با هوش مصنوعی در هوشا؛ از انتخاب مدل تا دریافت ویدیو

برای ساخت ویدیو با هوش مصنوعی در پلتفرم هوشا، ابتدا از یکی از ۳ طریق زیر سرویس ساخت ویدیو را انتخاب کنید:

سپس مراحل زیر را به‌ترتیب انجام دهید:

1. انتخاب مدل: مدل هوش مصنوعی موردنظر، مانند Grok Imagine 1.5 را انتخاب کنید.
2. بارگذاری تصویر: در صورت نیاز، عکس دلخواهتان را برای تبدیل به ویدیو آپلود کنید.
3. نوشتن پرامپت: موضوع ویدیو، حرکات، محیط و سبک تصویربرداری را توضیح دهید.
4. تنظیمات ویدیو: مدت زمان، کیفیت و نسبت تصویر را متناسب با نیازتان تنظیم کنید.
5. ساخت ویدیو: روی دکمه «تولید کن» کلیک کنید و منتظر بمانید تا ویدیو آماده شود. پس از آماده‌شدن، می‌توانید ویدیوی خود را دانلود کنید.

توجه: ویدیوهای ساخته شده در گالری هوشا برای ۶ ماه ذخیره می‌شوند.

راهنمای تولید ویدیو با ‪AI‬

۶ پرامپت ویدیو تولد آماده با قابلیت کپی و شخصی‌سازی

در ادامه ۶ پرامپت ساخت ویدیو تولد آماده برای کپی را برای‌تان به اشتراک گذاشته‌ایم. تنها کافی است در سرویس ساخت ویدیو هوشا، عکس یا عکس‌های خود را آپلود کنید، سپس پرامپت مورد نظر را کپی و جایگذاری کرده و در نهایت ویدیوی خود را بسازید.

همچنین می‌توانید برای ساخت پرامپت از صفر یا شخصی‌سازی آن از چت جی پی تی رایگان در هوشا نیز کمک بگیرید.

۱. ساخت کلیپ تولد لاکچری با هوش مصنوعی (با کاراکتر خانم)

با این پرامپت می‌توانید از عکس خود یک کلیپ تولد جذاب و متفاوت بسازید. در ویدیو، شخصیت اصلی با حفظ چهره و ظاهر عکس مرجع، در یک پارک شبانه با کیک تولد و بطری شامپاین دیده می‌شود. فوت کردن شمع‌ها، شعله‌های بزرگ آتش و باز شدن ناگهانی بطری شامپاین، لحظه‌ای غیرمنتظره و تماشایی را رقم می‌زند.

• ابعاد: ۹:۱۶
• مدل هوش مصنوعی: Grok Imagine 1.5

متن پرامپت: 

@image is the exact female character reference. The main character is a woman. Keep her exact same face, hairstyle, hair color, body build, skin tone, and exact clothing from @image in every frame. Do not change her face, hair, body, outfit, or accessories. She must remain recognizably the same woman throughout the entire video. Style: vertical smartphone video, 9:16, 1080×1920, handheld at night. Raw mobile footage, realistic shake, motion blur, low-light noise, street-lamp lighting only. Maximum photorealism, like a video filmed by a friend. Not cinematic. Duration: 10 seconds. Audio: night park ambience, traffic, wind, candle crackle. Climax: wooden “thunk”, loud cork POP, champagne fizz, fireball “WHOOOMP”. Soft female laughter at end. No speech, no music. Setting: ordinary nighttime city park. Dark sky, bare trees, warm lamps, dry ground, dead leaves, old wooden bench centered. Camera: handheld smartphone shot by an unseen friend. Organic shake, waist-to-head framing. Sequence:00:00-00:01.5: Empty bench. A realistic scene. A woman matching @image enters the frame from the left side of the screen and calmly walks toward a wooden bench in an outdoor setting, smiling silently. Her right hand holds a dark green champagne bottle vertically, cork up, base down. Her left hand carries a small black cake on a plate with tall burning orange candles. 00:01.5-00:03: She sits in the center of the bench, facing the camera and smiling. She extends her left hand holding the cake to her left, on the right side of the frame. Her right hand holds the champagne bottle on the opposite side. Candles burn brightly. 00:03-00:04: She shakes the champagne bottle vigorously with her right hand. The cake stays steady in her left hand. 00:04-00:05: Critical simultaneous action. The bottle stays vertical. She swings the bottle straight down and slams its flat bottom onto the bench seat on her right side, on the left side of the frame. No sideways swing, no rotation. At the same instant, she turns her head left toward the cake and blows flammable powder across the candles. The powder ignites into a huge realistic orange-yellow fireball, 1-1.5 meters wide. 00:05-00:08.5: Extreme slow motion, 1/8 speed. Two effects occur together:1. The cork shoots upward. A powerful champagne fountain erupts from the bottle, 1.5-2 meters tall, with clear-white and golden foam and droplets.2. A huge orange-yellow fireball blooms above the black cake, with billowing flames, dark smoke, embers, and natural sparks. No magic, fireworks, or artificial sparkles. The cake stays intact. She keeps the bottle vertical and the cake extended, with her head turned left, mouth open mid-blow, and eyes squinted. 00:08.5-00:10: Return to real time. The fireball fades into thin smoke. Champagne droplets fall onto her hair, shoulders, clothes, and the bench. The fountain weakens but keeps fizzing. She relaxes, still holding the bottle and cake, smiling and laughing softly at the camera. Champagne drips, with faint smoke around her face. Constraints: The character must always be the same woman from @image, with identical facial features, hairstyle, body proportions, skin tone, clothing, and accessories. No male character. No gender changes. No face morphing. No speech. Her right hand holds the champagne bottle and her left hand holds the black cake with candles throughout the entire clip. She must not drop either prop. Bottle strike must be vertical, with the flat base hitting the bench. Fire breath and bottle impact must happen simultaneously. Maximum photorealism and consistent female identity throughout.

۲. ساخت کلیپ تولد لاکچری با هوش مصنوعی (با کاراکتر آقا)

این کلیپ نیز خروجی مشابه با پرامپت قبلی دارد با این تفاوت که متناسب با تیپ و استایل آقایان نوشته شده است.

• ابعاد: ۹:۱۶
• مدل هوش مصنوعی: Grok Imagine 1.5

متن پرامپت:

@image is the exact character reference. Keep same face, hair, build, skin tone and exact clothing from @image every frame. Do not change face, hair, body, outfit or accessories. Style: vertical smartphone video, 9:16, 1080×1920, handheld at night. Raw mobile footage, realistic shake, motion blur, low-light noise, street-lamp lighting only. Maximum photorealism, like video filmed by a friend. Not cinematic. Duration: 10 seconds. Audio: night park ambience, traffic, wind, candle crackle. Climax: wooden “thunk”, loud cork POP, champagne fizz, fireball “WHOOOMP”. Soft laughter at end. No speech, no music. Setting: ordinary nighttime city park. Dark sky, bare trees, warm lamps, dry ground, dead leaves, old wooden bench centered. Camera: handheld smartphone shot by unseen friend. Organic shake, waist-to-head framing. Sequence:00:00-00:01.5: Empty bench. A cinematic, realistic scene. A man enters the frame from the left side of the screen and calmly walks toward a wooden bench in an outdoor setting, smiling silently. Right hand holds dark green champagne bottle vertical, cork up, base down. Left hand carries small black cake on plate with tall burning orange candles. 00:01.5-00:03: They sit center bench, facing camera, smiling. Left hand extends cake to their left, right side of frame. Right hand holds champagne opposite. Candles burn bright. 00:03-00:04: They shake champagne vigorously with right hand. Cake stays steady. 00:04-00:05: Critical simultaneous action. Bottle stays vertical. Person swings bottle straight down and slams flat bottom onto bench seat on their right side, left side of frame. No sideways swing, no rotation. At same instant, they turn head left toward cake and blow flammable powder across candles. Powder ignites into huge realistic orange-yellow fireball, 1-1.5 m wide 00:05-00:08.5: Extreme slow motion, 1/8 speed. Two effects together:1. Cork shoots up. Powerful champagne fountain erupts from bottle, 1.5-2 m tall, clear-white/golden foam and droplets.2. Huge orange-yellow fireball blooms above black cake, with billowing flame, dark smoke, embers and natural sparks. No magic, fireworks or sparkles. Cake stays intact. Person keeps bottle vertical, cake extended, head turned left, mouth open mid-blow, eyes squinted. 00:08.5-00:10: Real time. Fireball fades into thin smoke. Champagne droplets fall onto hair, shoulders, clothes and bench. Fountain weakens but keeps fizzing. Person relaxes, still holding bottle and cake, smiling and laughing softly at camera. Champagne drips, faint smoke around face. Constraints: same identity and clothing as @image every frame. No speech. Right hand holds champagne, left hand holds black cake with candles entire clip. Do not drop props. Bottle strike vertical, base down onto bench. Fire breath and bottle impact

۳. ساخت کلیپ تبریک تولد با هوش مصنوعی

با این پرامپت می‌توانید یک کلیپ تبریک تولد با فضایی متفاوت و نمایی از شهر نیویورک بسازید. در این ویدیو، دو نفر بالای ساختمان امپایر استیت ایستاده‌اند و پرچمی با پیام تبریک فارسی در دست دارند. دوربین با چند نمای متفاوت، از نمای نزدیک این دو نفر تا یک نمای هوایی از ساختمان و آسمان‌خراش‌های نیویورک، صحنه را ثبت می‌کند.

• ابعاد: ۹:۱۶
• مدل هوش مصنوعی: SeeDance 2 Pro

متن پرامپت:

جمله‌ی روی پرچم (باید دست‌نخورده بماند)» MULTI-SHOT VIDEO PROMPT (image-to-video, use the provided image  as the first frame; total duration ~12 seconds, 3 shots): GLOBAL RULES (apply to every shot): – Photorealistic aerial news-helicopter footage, documentary look,    desaturated grey-blue grade, gentle helicopter micro-vibration. – The couple, their black streetwear, the Empire State Building    and the Manhattan skyline stay EXACTLY consistent across all    shots — same people, same clothes, same weather, same light. – The black flag stays stretched and taut, rippling softly near    its trailing edge only. The Persian text on the flag remains    fully legible, undistorted and EXACTLY unchanged in every shot    and every frame. – No new objects, no new people, no text morphing, no fast camera    moves, natural motion blur only. SHOT 1 (0-4s) — CLOSE ON THE COUPLE: Camera level with the antenna tip, medium-close framing on the  couple and the flag filling the frame. The woman grips the mast,  her ponytail moving in the wind; the man holds the flag’s upper  corner, arm steady. The flag’s Persian text reads clearly, front  and center. Slow, almost imperceptible drift to the right. SHOT 2 (4-8s) — SLOW ORBIT: Cut to a slightly wider framing. The camera arcs smoothly from  left to right around the antenna tip, parallax between the spire,  the couple and the hazy cityscape far below. One World Trade  Center passes slowly in the deep background. The flag stays open  toward camera, text legible throughout the arc. SHOT 3 (8-12s) — PULL-BACK REVEAL: Cut to camera slightly above the antenna, looking gently down.  Steady slow pull-back revealing the full spire, the Art Deco  crown and the immense city dropping away below, streets and  rooftops receding into haze. The couple becomes small against the  skyline, the black flag with its Persian text remains the  strongest graphic element in frame. Ends on a wide, poster-like  composition.

۴. پرامپت ویدیو هوش مصنوعی تولد فرزند در کنار مادر؛ از کودکی تا بزرگسالی

با این پرامپت برای ویدیو تولد، یک کلیپ تولد احساسی بسازید که گذر زمان و عشق مادر و فرزند را به تصویر می‌کشد؛ از فوت کردن شمع‌ها در کودکی تا روشن کردن دوباره آن‌ها در بزرگسالی.

• ابعاد: ۱۶:۹
• مدل هوش مصنوعی: SeeDance 2.5

متن پرامپت:

SCENE CONTEXT A mother’s birthday told across two eras at the same table. A young mother holds her small son, they look at each other and blow out the candles together, dropping the room into darkness. In the dark a grown man strikes a match and relights the same cake, and the returning candlelight reveals an older woman sitting in the exact seat the young mother occupied, her grown son beside her. ACTIVE REFERENCES @YOUNGMOM: 30yo slim woman, long dark curly hair over her shoulders, navy sleeveless dress with small white polka dots, warm smile, both arms wrapped around the boy on her lap. 100% matches the reference. @BOY: 6yo boy, dense black curly hair, pale blue short-sleeve school shirt with chest pockets and a dark maroon tie, grey shorts, small hands gripping her shoulder. 100% matches the reference. @OLDMOM: 65yo woman, grey wavy chin-length hair, sage-green pintuck linen blouse, thin gold necklace, dark charcoal trousers, quiet closed-lip smile. 100% matches the reference. @SON: 30yo lean man, black curly hair, full dark moustache and stubble, unbuttoned oatmeal linen shirt with sleeves rolled to the forearm, black trousers, right hand holding a lit wooden match. 100% matches the reference. @CAKE: white buttercream sheet cake on a silver board, piped pink, lilac and cream rosettes with green leaves around the border, dark chocolate Persian script across the centre. Script shape, letters and layout match the reference exactly. No invented lettering, no extra text. LOCATION MAP A dim home dining room at night. Camera stands at the near table edge facing across it. Foreground: the cake centre-frame at the near edge, 0.6 meters from the lens, candle row at the top of the lower third. Midground: two people seated directly behind the cake, within 1 meter of it, torsos leaning toward the candles. Background: a dark cane-back dining chair screen-right, blurred furniture and a doorway falling to near-black. The candles are the only light source. Same room, same table, same camera position, same two seats in every shot. FIRST FRAME AND SPATIAL BLOCKING The first visible frame already contains @YOUNGMOM and @BOY behind the lit cake. No empty establishing frame, no delayed reveal. @YOUNGMOM sits screen-left of centre, chest 0.8 meters behind the cake, torso angled toward camera, both arms closed around the boy. @BOY sits on her lap screen-right of her, back against her chest, left arm hooked around her neck, his face within 15 centimetres of hers. Both heads turn inward: her eyes locked on his eyes, his eyes locked on hers. Neither looks at camera. The cake holds the lower centre third with the candle row unobstructed between lens and faces. FORMAT MODE CONTROLLED MULTI-SHOT SEQUENCE, three shots, real-time motion. HARD CUT at 0:07. MATCH CUT at 0:11. No fade, no crossfade, no dissolve, no transition effects, no invented extra cuts. No subtitles, no music, no dialogue. Lips stay closed except for the breath of blowing OPTICS SHOT A LENS LOCK = 84°. 84° diagonal field of view, classic wide-angle lens character, camera 1.2 meters from the cake, lens just above candle-tip height. Environment stays visible around the subjects to the frame edges, camera physically close, deep readable focus from candles to faces, straight table and chair lines stay rectilinear, no fisheye curve, no telephoto compression, no creamy portrait bokeh. SHOT B LENS LOCK = 29°. 29° diagonal field of view, short telephoto detail lens character, camera 1.2 meters from the match head, close framing achieved through lens reach. Razor focus on the burning match head, only the match and wicks are sharp, everything behind melts into a creamy dark bokeh wash, background compressed flat behind the flame. SHOT C LENS LOCK = 84°, identical to Shot A. Same camera position, height, distance and framing. Environment visible around the subjects, camera physically close, deep readable focus, rectilinear lines, no telephoto compression. No lens drift inside any shot. Lens character changes only at a cut. CAMERA Shot A: camera fixed at candle height, almost imperceptible settling drift, subjects in the upper two thirds, cake in the lower third. Shot B: camera fixed at the same table edge, framing tightened onto the match head and near candle row, focus rides the match head as it lowers. Shot C: camera returns to the exact Shot A position and framing so both eras occupy identical screen space. @OLDMOM’s head sits in the same screen position @YOUNGMOM’s head held. Slow focus settle from the flames back to their faces. ACTION TIMING 0:00 to 0:03 – SHOT A. Candles burn tall and steady. @YOUNGMOM tightens both arms around @BOY. He turns his head toward her, she turns toward him, their eyes meet. Both smile. Flame tips lean gently in the room air. 0:03 to 0:05 – @BOY presses his cheek toward her jaw in a hug, her chin lowers to the top of his head, both keep looking at each other, her hand strokes once across his shoulder. 0:05 to 0:07 – Both lean forward together over the cake, draw breath and blow. The flames bend hard away from camera, stretch thin and go out in an uneven ripple across the row. Grey smoke ribbons rise from every wick. The room falls to near-black within 0.4 seconds, only orange ember dots on the wicks and the faintest edge light on their shoulders. HARD CUT at 0:07. 0:07 to 0:08 – SHOT B. Frame opens already occupied: @SON’s right hand and a matchbox at screen-right, the dark cake edge and dead wicks readable low in frame, thin smoke still drifting. Near-black exposure, no empty frame. 0:08 to 0:09 – The match head scrapes the strike strip. A sharp sulfur flare punches out, then settles into a single tear-shaped flame. The match head is blackened and burnt, the carbon curling downward from the heat, the wooden stick charring behind it. 0:09 to 0:11 – He lowers the match to the cake and touches wick after wick. Each wick catches with a small pop of flame and light climbs step by step across the frame with every candle lit. MATCH CUT at 0:11. 0:11 to 0:13 – SHOT C. The full candle row burns. @OLDMOM sits in the exact seat, posture and screen position @YOUNGMOM held. @SON sits on her right where the boy sat, his right arm around her shoulders, the spent match still smoking in his lowered left hand. @CAKE is centre-frame, the Persian script clearly readable in the candlelight. 0:13 to 0:15 – She turns her head to him, he turns to her, eyes lock exactly as in Shot A. She lifts one hand to his forearm. Both hold the look, breathing, candles steady. Neither looks at camera. PHYSICS Flames respond to breath with real air pressure: they bend, elongate, detach and snuff in sequence, never simultaneously. Smoke rises in thin curling ribbons that drift and thin out. Wax pools and glistens at the candle tops. The struck match flares, drops in intensity, then stabilises; the burnt head curls under its own weight and sheds a flake of carbon. Hair moves a half-beat behind head turns. Fabric folds and settles with body weight. Hands make real contact pressure on shoulders and forearms with visible skin compression. No floating motion, no rubbery CG movement, no flames igniting or extinguishing all at once. LIGHTING Candlelight is the only source in every shot. It sits below and in front of the faces, throwing warm uplight on cheeks, chins and the underside of the brow, tops of heads falling into shadow, background dropping to black. Exposure is set for the flames, not the faces. Light level tracks the action exactly: full glow at the start of Shot A, near-total darkness from 0:06 to 0:08, a stepped rise back to full glow across 0:09 to 0:11. No fill light, no practical lamps, no flat front key. SHOT A GRADE: warm nostalgic 35mm film emulation in the character of Kodak Portra 400. Temperature pushed warm, deep saturated golden-yellows anchoring the frame, soft creamy highlight roll-off with no clipped whites, deeper reds slightly desaturated into an analog profile while still holding contrast against the dark, shadows lifted with gentle warm brown undertones and never true black, light haze in the air, fine organic film grain. SHOT B AND SHOT C GRADE: clean modern digital cinema. Neutral-warm candle colour with accurate skin tones, crisp sharp glass, high micro-contrast, true clean blacks, no haze, no grain, no vintage softness, no lifted-shadow film wash. AUDIO Quiet room tone. Cloth rustle and two breaths drawn before the blow, a soft double exhale, then near-silence with a faint smoke hiss. The scrape and crack of the match against the box, a brief sulfur hiss, small flame catches as each wick lights. Faint breathing under the last beat. No music, no voices, no subtitles. POSITIVE CONSTRAINTS Exactly two people per shot: @YOUNGMOM and @BOY in Shot A, @SON’s hands in Shot B, @OLDMOM and @SON in Shot C. No extra guests, no duplicate characters, no stray hands at the frame edges. @OLDMOM occupies the identical seat, posture and screen position as @YOUNGMOM. @SON occupies the identical side as @BOY. No swapped sides, no reversed screen direction. Gaze stays locked between the two subjects in Shot A and Shot C. No one looks into the lens. The Persian script on @CAKE matches the reference exactly. No garbled letters, no added words, no Latin text. Faces are lit only by candlelight from below. No flat front lighting, no beauty fill. Sharp clarity, natural colour, stable picture, no blur, no ghosting, no flickering outside the candle flames themselves.

۵. ساخت کلیپ تولد با هوش مصنوعی انلاین در فضا

با این پرامپت می‌توانید یک کلیپ تولد با حال‌وهوای علمی‌تخیلی بسازید. در این سناریو، دو نفر با لباس فضانوردی در یک سفینه فضایی و در مقابل منظره زمین شناور هستند. کیک تولد هم داخل یک محفظه شیشه‌ای بین آن‌ها قرار دارد و با حرکت آرام در فضا جابه‌جا می‌شود. تمام جزئیات، از چهره و ظاهر دو نفر گرفته تا طراحی کیک و فضای سفینه، بر اساس تصاویر مرجع و تنظیمات پرامپت ساخته می‌شوند.

• ابعاد: ۱۶:۹
• مدل هوش مصنوعی: SeeDance 2 Pro

پرامپت:

=== INPUT PARAMETERS === REF_WOMAN = @[Image 1](image_1)   → Image 1 is ALWAYS the woman (the birthday person) REF_MAN   = @[Image 2](image_2)   → Image 2 is ALWAYS the man (the partner) WOMAN_DESC = [fill in: age, build, hair length + texture + how it is worn, eye colour, one or two unique features] MAN_DESC   = [fill in: age, build, hair length + texture + how it is worn, facial hair or bare face, one or two unique features] WORLD_SOURCE = [film title for this world, or leave empty for the default world] ASPECT = 16:9 DURATION = 12 seconds === END INPUT PARAMETERS === SCENE CONTEXT World four. Default world: the observation deck of a sleek spacecraft, Earth filling the panoramic window. The couple floats weightless in flight suits, the cake drifts between them inside a glass dome, its flames burning as small glowing spheres. Playful zero gravity joy. The couple are together in frame through the whole scene. CHARACTER LOCK Scan every physical trait from the attached references and reproduce it exactly, never hardcode, never idealise, never stylise, never age, never re-proportion. @[Image 1](image_1) is the woman per WOMAN_DESC, @[Image 2](image_2) is the man per MAN_DESC. Face geometry, bone structure, skin tone, eye shape and colour, brows, nose, lips, hairline, hair colour and texture and facial hair exactly as in the references. Identities never blended, never swapped. Faces bare and eyes fully visible in every shot, no eyewear, no mask, no face covering. Pore-level skin realism, real catch-lights, natural blinking. ATMOSPHERE CAST LOCK Every background person is an invented face, randomised by the model and different on every run. No person in any frame may resemble any real actor, celebrity, public figure or named film character, and no costume may read as one specific named character. No logos, no brand marks, no readable text anywhere in frame. CAKE LOCK One single birthday cake design carried through the whole film: a small round cake with smooth white cream, one thin gold piped line around the base, and lit candles. Identical shape, size, colour and candle layout in every world it appears in. It is the only cake in any frame. WORLD SOURCE WORLD_SOURCE holds one film title given by the customer. Translate that film’s world using only location type, architecture, lighting, colour palette and wardrobe era. Never depict its named characters, any actor’s likeness, its logos, its unique props or any readable title. If WORLD_SOURCE is empty, use the default world described in this prompt exactly as written. LOCATION MAP A clean white-and-graphite observation deck, one vast curved window with Earth below in daylight blue, console light strips in cool white, two invented-face crew members strapped at stations in the background. No readable displays. The cake floats mid-deck inside a clear glass dome. White balance 5600K with cool console accents. FIRST FRAME / BLOCKING First frame: WS across the deck, Earth filling the window, the couple already floating mid-air in slow drift, hair lifting weightless, the glass-domed cake rotating slowly between them, candle flames burning as round golden spheres. FORMAT MODE Timed multishot, 12 seconds total, three segments. Cuts only at 4.0s and 8.0s, the camera does not cut on its own. OPTICS 0.0s to 4.0s: WS, 63°. 4.0s to 8.0s: MCU, 29°, shallow depth of field. 8.0s to 12.0s: MS two-shot, 47°. No drift mid-segment. Rectilinear, natural motion blur. CAMERA 0.0s to 4.0s: locked off, the drift of the subjects supplying all motion. 4.0s to 8.0s: slow float-in at 1 km/h toward the dome and their faces. 8.0s to 12.0s: locked off with Earth centred behind them. Wide latitude, window highlights protected. ACTION 0.0s to 4.0s — The couple drifts weightless, she upside down relative to him, both laughing silently. The domed cake rotates between them, sphere-flames steady inside. 4.0s HARD CUT 4.0s to 8.0s — She nudges the dome with two fingers and it glides gently to him across frame, flames wobbling like liquid gold. He catches it, spins it once, the reflection of Earth sliding over the glass. 8.0s HARD CUT 8.0s to 12.0s — He pushes off gently and floats to her, they align right side up holding the dome together between them, Earth huge and slow behind, hair and suit straps drifting. PERFORMANCE Weightless play: silent laughter, exaggerated slow-motion-feeling movement that is really just real zero gravity inertia, one held look at the end with Earth behind. No dialogue. PHYSICS True zero gravity: every push has an equal drift, nothing falls, hair and straps float with inertia, the dome travels in a straight line until caught, flames behave as glowing spheres that wobble when the dome is nudged. Console lights reflect on the glass. LIGHTING Earth-light through the window as the broad cool key, white deck bounce as fill, warm accent only from the sphere-flames inside the dome. Clean, bright, spacious. COLOR GRADE Deep space black and Earth blues against clean white deck, the warm gold of the flames and skin tones as the only warmth. Crisp and open, the coolest palette of the five worlds. WARDROBE Both leads: fitted flight suits in off-white and graphite with no insignia, no helmets, faces and hair fully visible and floating. Crew: same suit family, invented faces. RANDOMISATION Console light patterns kept text-free, crew faces, deck props and Earth’s cloud layout are invented by the model, different on every run. WORLD_SOURCE, when set, replaces this default world per the WORLD SOURCE block. AUDIO Soft ship hum, air system whisper, a muffled happy laugh through held breath, one clean glass tap as the dome is caught. No dialogue. STYLE Photoreal live-action cinema, 8K detail, fine natural film grain, natural motion blur, real-time speed in every segment. OUTPUT SETTINGS 12 seconds, 16:9, 8K photoreal, real-time speed in every segment. POSITIVE LOCKS @[Image 1](image_1) is the woman and @[Image 2](image_2) is the man, never reversed, both identical to the references and together in every segment, no helmet ever covering a face. The cake matches CAKE LOCK exactly, inside its glass dome in every shot of this scene. No readable text on any display. Cuts only at 4.0s and 8.0s. HANDOFF FRAME Last frame: Earth’s night side edging into frame behind them, city lights glittering. The scattered city lights match-cut into rain-blurred neon bokeh that opens T1-P6

۶. ساخت کلیپ تولد در هوش مصنوعی با حال‌وهوای جادویی و فانتزی

با این پرامپت می‌توانید یک کلیپ تولد با فضایی شبیه تالارهای جادویی فیلم‌های فانتزی بسازید. در این ویدیو، زوج اصلی در یک تالار بزرگ و قدیمی با لباس‌های مخصوص آن فضا حضور دارند. شمع‌های شناور در هوا و نور گرم آن‌ها، فضای متفاوتی برای جشن تولد ایجاد می‌کنند. بعد از خاموش کردن شمع‌های کیک، ذرات نورانی طلایی در فضای تالار پخش می‌شوند و صحنه حال‌وهوایی جادویی پیدا می‌کند. چهره و ظاهر افراد نیز مطابق تصاویر مرجع حفظ می‌شود.

• ابعاد: ۱۶:۹
• مدل هوش مصنوعی: SeeDance 2 Pro

متن پرامپت:

=== INPUT PARAMETERS === REF_WOMAN = @[Image 1](image_1)   → Image 1 is ALWAYS the woman (the birthday person) REF_MAN   = @[Image 2](image_2)   → Image 2 is ALWAYS the man (the partner) WOMAN_DESC = [fill in: age, build, hair length + texture + how it is worn, eye colour, one or two unique features] MAN_DESC   = [fill in: age, build, hair length + texture + how it is worn, facial hair or bare face, one or two unique features] WORLD_SOURCE = [film title for this world, or leave empty for the default world] ASPECT = 16:9 DURATION = 15 seconds === END INPUT PARAMETERS === SCENE CONTEXT World two. Default world: a vast medieval stone great hall at night where hundreds of small candles float in the air, long banquet tables, robed invented-face students. The couple sits at the high table with the cake already lit in front of them. She blows the candles out ONCE, and everything after that is the quiet magic of the hall itself. Wonder and quiet magic. The couple are together in frame through the whole scene. CHARACTER LOCK Scan every physical trait from the attached references and reproduce it exactly, never hardcode, never idealise, never stylise, never age, never re-proportion. @[Image 1](image_1) is the woman per WOMAN_DESC, @[Image 2](image_2) is the man per MAN_DESC. Face geometry, bone structure, skin tone, eye shape and colour, brows, nose, lips, hairline, hair colour and texture and facial hair exactly as in the references. Identities never blended, never swapped. Faces bare and eyes fully visible in every shot, no eyewear, no mask, no face covering. Pore-level skin realism, real catch-lights, natural blinking. ATMOSPHERE CAST LOCK Every background person is an invented face, randomised by the model and different on every run. No person in any frame may resemble any real actor, celebrity, public figure or named film character, and no costume may read as one specific named character. No logos, no brand marks, no readable text anywhere in frame. CAKE LOCK One single birthday cake design carried through the whole film: a small round cake with smooth white cream, one thin gold piped line around the base, and thin candles. Identical shape, size, colour and candle layout in every world it appears in. Candle state is NOT part of this lock: it follows the ACTION timeline exactly. It is the only cake in any frame. WORLD SOURCE WORLD_SOURCE holds one film title given by the customer. Translate that film’s world using only location type, architecture, lighting, colour palette and wardrobe era. Never depict its named characters, any actor’s likeness, its logos, its unique props or any readable title. If WORLD_SOURCE is empty, use the default world described in this prompt exactly as written. LOCATION MAP Gothic stone hall with tall arched windows showing a starry night, four long wooden tables of robed figures, an open ceiling that reads as night sky, hundreds of floating candles hanging at different heights as the dominant light. The high table on a stone dais. White balance 3400K, fixed. FIRST FRAME / BLOCKING First frame: WS from the back of the hall down the central aisle, floating candle field filling the upper half, the couple small at the high table in dark scholars’ robes, the white cake in front of them, its candles ALREADY LIT and burning steadily. No candle is ever ignited on screen at any point in this film. FORMAT MODE Timed multishot, 15 seconds total, four segments. Cuts only at 4.0s, 9.0s and 12.0s, the camera does not cut on its own. OPTICS 0.0s to 4.0s: WS, 63°. 4.0s to 9.0s: MCU two-shot, 29°, shallow depth of field. 9.0s to 12.0s: MS, 47°. 12.0s to 15.0s: WS, 63°. No drift mid-segment. Rectilinear, soft round bokeh on the candle field. CAMERA 0.0s to 4.0s: slow glide forward down the aisle at 3 km/h, 2 metres above the floor, rising gently. 4.0s to 9.0s: locked off across the high table. 9.0s to 12.0s: slow orbit a quarter turn around the couple at 1 km/h. 12.0s to 15.0s: slow rising pull-back away from the dais down the aisle at 3 km/h, tilting gently up into the floating candle field. Wide latitude, every flame holds detail. ACTION 0.0s to 4.0s — The glide carries through the hall between the floating flames, robed figures turning to watch the couple ahead at the high table, the small cake flames already burning steadily in front of them. Candlelight ripples across stone. 4.0s HARD CUT 4.0s to 9.0s — Locked two-shot across the cake, its candles already lit and burning calmly. She smiles at him, closes her eyes for a silent wish, and at 7.0s she blows out the candles with ONE single exhale: every flame on the cake dies instantly with her breath and stays out for the rest of the film, thin grey wisps of smoke curling up from the wicks, the warm under-light fading from their faces. This is the ONLY blow in the entire film. After it she opens her eyes with a small laugh; she never leans toward the cake or breathes toward it again. 9.0s HARD CUT 9.0s to 12.0s — Fine golden sparks of light drift down from the floating field like slow snow around the couple, catching in her hair and on his shoulders. The smoking cake stays extinguished. The hall of robed figures raises cups in a silent toast while the couple look up at the falling sparks together. 12.0s HARD CUT 12.0s to 15.0s — Pure atmosphere, no new events: the rising pull-back reveals the whole hall under the floating candle field, golden sparks falling everywhere through the air, the couple small and warm at the high table, the starry ceiling deep above. The cake candles remain out, thin smoke long dissolved. PERFORMANCE Pure wonder on both: parted lips, slow blinks, her eyes closing for the silent wish, one single controlled exhale at 7.0s and never a second one, her small laugh right after the flames die, his eyes on her more than on the magic. No dialogue. PHYSICS Floating candles sway very slightly on air currents, never loop. When she blows, the cake flames bend away from her breath and extinguish instantly and completely, no flame survives, none relights, thin smoke rises and dissolves, each wick keeps a brief orange ember glow for about one second. Breath can ONLY extinguish flame, it can never ignite a candle; no flame ignites on screen anywhere in this film. Only the cake candles go out; the floating hall candles stay lit. The golden sparks fall with light air resistance and glow out on contact. Robes hang heavy. Contact shadows anchor everyone. LIGHTING The floating candle field is the only source: warm top-light from hundreds of small flames, soft pools on the tables, faces underlit by the burning cake flames from 4.0s until the blow at 7.0s. After the blow-out the under-light fades with the flames, leaving only the warm top-light and the falling sparks. Starry ceiling stays deep. COLOR GRADE Warm candle gold against deep blue-black stone and night ceiling, dark robe blacks kept clean, cream cake bright at centre. High wonder, no harshness. WARDROBE Both leads: floor-length dark scholars’ robes over simple period clothes, faces and hair fully visible per the references. Students: identical style robes, invented faces, muted house-neutral colours with no emblems. RANDOMISATION Every student’s face, banner-free wall details, table props and the exact candle field layout are invented by the model, different on every run. WORLD_SOURCE, when set, replaces this default world per the WORLD SOURCE block. AUDIO Vast quiet stone ambience, hundreds of tiny flames flickering, a soft choral hum far away, one soft breath and the faint puff of all the small flames dying together at 7.0s, then a delicate magical shimmer as the golden sparks fall. No dialogue. STYLE Photoreal live-action cinema, 8K detail, fine natural film grain, natural motion blur, real-time speed in every segment. OUTPUT SETTINGS 15 seconds, 16:9, 8K photoreal, real-time speed in every segment. POSITIVE LOCKS @[Image 1](image_1) is the woman and @[Image 2](image_2) is the man, never reversed, both identical to the references and together in every segment. The cake matches CAKE LOCK exactly. The cake candles are already burning at the first frame and nothing ever lights a candle on screen. There is exactly ONE blow in the whole film, at 7.0s; blowing only extinguishes and never ignites. From 7.0s to the final frame the cake candles remain smoking and unlit, they never relight and never flicker back; the floating hall candles never go out. No emblem, crest or readable text anywhere, no background face resembling any real actor or named character. Cuts only at 4.0s, 9.0s and 12.0s. HANDOFF FRAME Last frame: the wide hall with golden sparks drifting through the whole frame. The drifting gold specks match-cut into backlit golden dust of the saloon that opens T1-P4

ساخت کلیپ تولد با هوش مصنوعی اندروید

هوشا علاوه بر سایت هوش مصنوعی، نسخه اندروید هم دارد و می‌توانید از طریق اپلیکیشن آن، مراحل ساخت ویدیو را با گوشی انجام دهید. تمامی مراحل ساخت ویدیو در نسخه اندروید مشابه نسخه مرورگر است، شما تنها باید عکس خود را آپلود کنید، پرامپت را وارد کنید و پس از تنظیم خروجی، ویدیوی خود را بسازید.

هر ۶ پرامپت بالا را می‌توانید در گوشی موبایل خود نیز امتحان کنید (تمامی پرامپت‌ها از بخش استایل‌های آماده هوشا هم در نسخه وب و هم نسخه اندروید نیز در دسترس هستند).

دانلود اپلیکیشن هوشا برای اندروید؛ گوگل پلی، بازار و لینک مستقیم

ساخت آهنگ تولد با هوش مصنوعی برای ویدیوی شخصی

شما می‌توانید با کمک هوش مصنوعی آنلاین، موزیک کلیپ خودتان را هم بسازید. ویدیو فقط نیمی از حس کلیپ تولد است اما موسیقی خوب می‌تواند یک ویدیوی ساده را خیلی شخصی‌تر کند.

هوشا در کنار سرویس‌های ساخت ویدیو و ساخت عکس، بخش ساخت موسیقی هم دارد. در این سرویس شما می‌توانید یک توضیح کوتاه درباره آهنگ سبک موردنظر و فضای آن بدهید و موسیقی مخصوص کلیپ تان را بسازید.

آموزش ساخت آهنگ تولد اختصاصی برای کلیپ با هوش مصنوعی

برای شروع ساخت آهنگ با هوش مصنوعی به یکی از ۳ روش زیر، صفحه سرویس مربوط را در هوشا باز کنید:

حال طبق مراحل زیر پیش بروید:

1. افزودن شعر: با فعال کردن این گزینه، می‌توانید متن شعر دلخواه خود را برای ساخت آهنگ وارد کنید.
2. بی‌کلام: با فعال کردن این گزینه، موسیقی بدون صدای خواننده ساخته می‌شود.
3. توضیحات موزیک: سبک، حال‌وهوا و ویژگی‌های موسیقی موردنظرتان را توصیف کنید. مثلا: «یک آهنگ کوتاه و شاد برای تولد سارا بساز. فضای آهنگ صمیمی و خانوادگی باشد، ریتم پاپ ملایم داشته باشد و اسم سارا در بخش اصلی آهنگ گفته شود.»
4. متن موزیک: متن ترانه‌ای را که می‌خواهید در آهنگ خوانده شود، وارد کنید.
5. تولید کن: پس از تکمیل تنظیمات، روی این دکمه کلیک کنید تا هوش مصنوعی موسیقی شما را بسازد.

معرفی ۶ هوش مصنوعی ساخت کلیپ تولد با عکس

برای ساخت کلیپ تولد با هوش مصنوعی رایگان و اشتراکی، ابزارهای مختلفی وجود دارند که می‌توانند عکس‌های شخصی را به ویدیو تبدیل کنند، افکت‌های جذاب بسازند یا چند تصویر و ویدیو را در قالب یک کلیپ کنار هم قرار دهند. در ادامه، چند برنامه ساخت کلیپ تولد با هوش مصنوعی را معرفی می‌کنیم.

۱. گوگل فلو (Google Flow)

گوگل فلو یکی از گزینه‌های مناسب برای ساخت کلیپ‌های تولد واقع‌گرایانه و خلاقانه است. با استفاده از این ابزار می‌توانید از متن و تصاویر مرجع برای تولید ویدیو کمک بگیرید و صحنه‌هایی با جلوه‌های ویژه، حرکت دوربین و فضای سینمایی بسازید. برای مثال، می‌توانید عکس خود را به یک صحنه تولد با کیک، شمع و افکت‌های آتش تبدیل کنید.

معرفی و آموزش هوش مصنوعی ساخت ویدیو گوگل فلو

۲. گوگل ویدز (Google Vids)

گوگل ویدز بیشتر برای ساخت و تدوین ویدیو کاربرد دارد. با این ابزار می‌توانید عکس‌ها و ویدیوهای شخصی را کنار هم قرار دهید، از قابلیت‌های هوش مصنوعی برای تولید محتوای ویدیویی کمک بگیرید و با افزودن موسیقی، متن و افکت، یک کلیپ تبریک تولد بسازید. این ابزار برای افرادی مناسب است که می‌خواهند بدون تدوین حرفه‌ای، ویدیویی مرتب و جذاب آماده کنند.

معرفی و آموزش هوش مصنوعی ساخت ویدیو گوگل ویدز

۳. جمینای اومنی (Gemini Omni)

Gemini Omni یک مدل هوش مصنوعی برای تولید و ویرایش محتوای ویدیویی است. با استفاده از قابلیت‌های آن می‌توانید از تصاویر مرجع برای ساخت ویدیو کمک بگیرید و صحنه‌های مختلفی را با دستورهای متنی ایجاد کنید. این مدل برای ساخت کلیپ‌های شخصی‌سازی‌شده که در آن‌ها حفظ ظاهر افراد و هماهنگی صحنه‌ها اهمیت دارد، کاربرد دارد.

معرفی و آموزش هوش مصنوعی ساخت ویدیو جمنای اومنی

۴. پیکا (Pika)

پیکا یکی از ابزارهای تولید ویدیو با هوش مصنوعی است که امکان تبدیل عکس به ویدیو و اضافه کردن جلوه‌های تصویری را فراهم می‌کند. قابلیت‌هایی مانند Pikaffects به شما اجازه می‌دهند افکت‌های خلاقانه‌ای به تصاویر اضافه کنید. برای مثال، می‌توانید عکس خود را به ویدیویی متحرک با جلوه‌های فانتزی و متفاوت تبدیل کنید.

معرفی و آموزش هوش مصنوعی ساخت ویدیو پیکا

۵. مجیستو (Magisto)

مجیستو بیشتر برای تدوین خودکار و ساخت کلیپ از عکس‌ها و ویدیوهای شخصی طراحی شده است. با انتخاب تصاویر، موسیقی و قالب مورد نظر می‌توانید یک ویدیوی خاطره‌انگیز برای تولد بسازید. این ابزار برای ساخت کلیپ‌های ساده و صمیمی مناسب‌تر است و امکانات آن با ابزارهای تولید ویدیو از پرامپت تفاوت دارد.

معرفی و آموزش هوش مصنوعی ساخت ویدیو مجیستو

۶. جنمو (Genmo)

جنمو یکی دیگر از ابزارهای هوش مصنوعی در حوزه تولید ویدیو است که با مدل‌هایی مانند Mochi امکان ساخت ویدیو از توضیحات متنی را فراهم می‌کند. با کمک آن می‌توانید ایده‌های مختلفی را به صحنه‌های متحرک تبدیل کنید. البته این ابزار بیشتر برای آزمایش و تولید ویدیوهای خلاقانه کاربرد دارد و برای ساخت کلیپ تولد از تصاویر شخصی، گزینه‌های دیگری مثل گوگل فلو و پیکا انتخاب‌های مناسب‌تری هستند.

معرفی و آموزش هوش مصنوعی ساخت ویدیو جنمو

جمع‌بندی‌

برای ساخت کلیپ تولد با هوش مصنوعی، بهتر است با یک عکس واضح و پرامپتی ساده شروع کنید. حرکت‌های طبیعی و حفظ شباهت چهره، از اضافه کردن افکت‌های متعدد اهمیت بیشتری دارند. برای این کار شما نیاز به مهارت خاصی ندارید و می‌توانید به‌راحتی و با استفاده از قالب‌های آماده هوشا، بدون طراحی صحنه از صفر، ایده خود را به یک ویدیوی شخصی‌سازی شده تبدیل کنید.

نویسنده: زهرا کاظمی