آموزش هوش مصنوعی | ابزار و پرامپت تولید محتوا
PolloAI_MiniMax_H3_Cyberpunk_Video_Tutorial_FA.pdf
569.6 KB
❤10👍2
Use the attached face image as the exact identity reference
and generate the same person.
IDENTITY LOCK — highest priority:
Preserve the exact facial identity, facial proportions, face shape,
eyes, eyebrows, nose, lips, jawline, cheekbones, ears, skin tone,
facial hair, tattoos and all natural facial characteristics from
the uploaded reference. Do not redesign, beautify or reinterpret
the face.
HAIR LOCK:
Preserve exactly the natural hairstyle visible in the uploaded
reference image — identical hairline, density, volume, length,
texture, shape and natural direction. Never restyle, shorten,
straighten, enlarge or alter the hair.
Create a highly realistic casual smartphone photograph on a
narrow wet forest road immediately after rain.
A glossy black BMW sedan is parked on the left side of the frame,
seen from a near rear-left three-quarter angle. The rear section
of the BMW is clearly visible, including the trunk, rear windshield,
left taillight, rear bumper, rear wheel and part of the left side of
the vehicle.
The BMW body, roof, rear windshield and trunk are covered with
small realistic rain droplets, with natural reflections of the
surrounding trees and cloudy sky on the wet black paint.
A small brown takeaway coffee cup with a black lid is standing
upright on top of the trunk near the rear windshield.
The person stands beside the BMW on the right side of the frame,
approximately beside the rear passenger-door/trunk area.
He is shown full body.
He wears a plain oversized black pullover hoodie, loose black
sweatpants and black low-top sneakers.
His torso is facing slightly toward the BMW while his head is turned
to his right, looking naturally off-camera toward the forest.
His right hand is raised casually to the upper chest, lightly holding
or adjusting the hoodie near the collar/drawstring.
His left arm hangs naturally relaxed beside his body.
Maintain a calm, masculine, thoughtful and spontaneous posture,
as if the photograph was captured naturally during a short stop on
the road.
The road is narrow, dark and completely wet from recent rain, with a pale
road-edge line running close beside the person. Wet brown leaves,
branches and forest debris cover the roadside.
Surround the scene with tall dense trees, eucalyptus-like trunks, deep green
vegetation and subtle natural mist between the trees. The road disappears
softly into fog in the distance.
Lighting should be completely natural and overcast, with soft diffused
daylight filtering through the cloudy sky and forest canopy. Cool neutral
color temperature, subdued greens and grays, realistic post-rain atmosphere.
Capture the image with a modern smartphone main camera, approximately
24–26mm equivalent lens, from around chest height and roughly 2.5 meters away.
The photograph should feel slightly imperfect and spontaneous: realistic
smartphone HDR, natural sharpening, subtle digital noise, moderate depth of field,
realistic exposure and no professional cinematic lighting.
The final result should look like a genuine casual smartphone photo taken by another
person during a rainy road trip — not a fashion campaign and not an automotive
advertisement.
Vertical 4:5 composition.
Negative prompt:
different person, altered identity, altered face, changed facial proportions, different
hairstyle, modified hairline, bald, shaved, receding hairline, thinning hair, buzz cut,
altered hair texture, artificial model pose, exaggerated muscular body, beauty filter,
plastic skin, skin smoothing, CGI, 3D render, cartoon, illustration, fake rain, excessive
fog, fantasy atmosphere, luxury automotive advertisement, studio lighting, cinematic
movie lighting, warped BMW, distorted vehicle geometry, malformed wheels,
false reflections, excessive bokeh, extra limbs, extra fingers, malformed hands, low quality.همراه ما باشید. 🙏🏻
@ai_farshad
Please open Telegram to view this post
VIEW IN TELEGRAM
😁17❤10👍4
Ultra-realistic podcast studio portrait, [MANTER EXATAMENTE O ROSTO DA FOTO DE REFERÊNCIA], seated indoors behind a desk microphone, black hoodie, hands together on the table, direct eye contact, calm confident expression. Dark cozy studio with warm hanging lights and plants in the background, cinematic lighting, shallow depth of field, crisp facial detail, realistic skin texture, professional creator setup, premium editorial realism.
Negative prompt: distorted face, extra fingers, blur, low detail, CGI, cartoon.همراه ما باشید. 🙏🏻
@ai_farshad
Please open Telegram to view this post
VIEW IN TELEGRAM
❤15👍6😁3
Use my uploaded photo as the exact identity reference. Preserve my facial identity, facial structure, skin tone, eyes, nose, lips, and natural features. Do not change who I am.
Create a photorealistic lifestyle portrait inspired by the reference image.
A young woman sitting casually indoors on the floor, holding a natural wood acoustic guitar across her lap. She is wearing a simple loose white short-sleeve T-shirt and relaxed blue jeans.
Pose: seated with one knee slightly raised, guitar resting naturally across the thighs. One hand is gently placed over the guitar strings near the sound hole, while the other hand holds the guitar neck naturally. Her shoulders are relaxed. Her head is slightly tilted downward and she is looking softly toward the guitar with her eyes partially lowered, creating a peaceful, intimate and candid expression.
Long natural hair falling loosely around the shoulders and over the guitar, slightly messy and effortless. Minimal natural makeup, soft lips, subtle blush, realistic skin texture.
Background: a simple plain light blue-gray wall and a quiet indoor environment with no distracting objects. Warm natural ambient light mixed with soft window light, gentle shadows, cozy evening atmosphere.
The guitar should be a realistic full-size warm honey-colored wooden acoustic guitar with visible wood grain, strings, bridge, sound hole and realistic reflections.
Photography style: authentic casual lifestyle photography, slightly vintage digital-camera aesthetic, soft warm tones, subtle film grain, gentle contrast, slightly muted colors, natural skin tones, soft highlights, realistic shadows, candid composition.
Camera positioned approximately at eye/chest level, vertical portrait composition, medium shot, 50mm lens look, shallow but natural depth of field. Keep the subject and guitar sharply detailed while the background is slightly soft.
Highly realistic photography, natural anatomy, realistic hands and fingers, realistic guitar proportions, authentic candid moment, no studio glamour, no artificial beauty filter.همراه ما باشید. 🙏🏻
@ai_farshad
Please open Telegram to view this post
VIEW IN TELEGRAM
❤18👍5😁2😍1
✓ راهنمای ۲۵ صفحهای؛ از ساخت تصویر تا تدوین نهایی
✓ پرامپتهای آماده و کپیکردنی برای هر ۵ سبک
✓ تمرین سهشاتی + راهکار رفع خطاهای رایج
✓ ضمیمه ۱۴ صفحهای آموزش Google Flow
همراه ما باشید. 🙏🏻
@ai_farshad
Please open Telegram to view this post
VIEW IN TELEGRAM
❤13🤬3👍2😢1
🚨 خبر مهم برای علاقهمندان به AI
🔵 مدل جدید و پرسروصدای Jev در دسترس قرار گرفته و فعلاً میتونید رایگان ازش استفاده کنید! 🔥
🔵 هر اکانت ۵ دلار اعتبار اولیه دریافت میکنه که میتونید برای استفاده از مدل مصرفش کنید.
🔵 برای شروع از این لینک وارد شوید:
🔗 https://console.typesafe.ai
همراه ما باشید. 🙏🏻
@ai_farshad
همراه ما باشید. 🙏🏻
@ai_farshad
Please open Telegram to view this post
VIEW IN TELEGRAM
❤22
گراک 4.7 با یک مدل پایه بزرگتر و آموزش طولانیتر برای حل مسائل دشوار توسعه داده شده و توانایی بهتری در بررسی و اصلاح خروجیهای خودش دارد.
این مدل برای کار با کد، اسناد، تولید محتوا و وظایف مهندسی و دانشمحور بهبود پیدا کرده است.
یعنی میتواند حجم بسیار بزرگی از متن و اطلاعات را در یک درخواست پردازش کند.
گراک 4.7 در بعضی آزمونهای تخصصی عملکرد بسیار خوبی داشته، اما در برخی تستها مدلهایی مثل GPT-5.6 Sol و Fable 5.1 امتیاز بالاتری گرفتهاند؛ بنابراین عملکرد آن به نوع کار و بنچمارک بستگی دارد.
@ai_farshad
Please open Telegram to view this post
VIEW IN TELEGRAM
❤14
Use the attached face image as the exact identity reference and generate the same person. Preserve the exact facial identity, facial proportions, skin tone, facial hair, and natural features from the uploaded reference.
IMPORTANT:
Preserve exactly the natural hair shown in the reference image — same hairstyle, hairline, density, volume, length, and texture. Never change the natural hairstyle.
Create a chiaroscuro lighting portrait with strong dramatic contrast between light and shadow. The subject is photographed in a dark environment, with one side of the face illuminated and the other side falling into deep shadow. The expression should be intense, confident, and slightly mysterious.
Frame the image as a close-up portrait with high facial detail, cinematic contrast, and a premium editorial feel. Use realistic skin texture and natural shadow transitions. Shot on a 105mm lens.
Cinematic, photorealistic, 8k, no text, no graphic elements, no watermark.همراه ما باشید. 🙏🏻
@ai_farshad
Please open Telegram to view this post
VIEW IN TELEGRAM
❤15👍3😁3
تجربه تدوین حرفهای با نسخه PRO کپکات
✨ بدون محدودیت
⚡️ فعالسازی سریع روی iOS
ضمانت اکانت | کیفیت خروجی حرفهای
━━━━━━━━━━━━━━
➖ یکماهه 750.000 تومن➖ سهماهه 1.100.000 تومن➖ ششماهه 2.000.000 تومن➖ یکساله 3.400.000 تومن
━━━━━━━━━━━━━━
✓ دسترسی کامل به افکتها و ترنزیشنهای حرفهای
✓ حذف واترمارک
✓ خروجی با کیفیت بالا
✓ مناسب تولیدکنندگان محتوا و ادیتورهای حرفهای
━━━━━━━━━━━━━━
Please open Telegram to view this post
VIEW IN TELEGRAM
❤7👍1😁1
This media is not supported in your browser
VIEW IN TELEGRAM
✓ ساخت Character Sheet و محیط
✓ آمادهسازی و تگگذاری رفرنسها
✓ طراحی یک سکانس ۸ ثانیهای در ۴ شات
✓ نوشتن و استفاده از پرامپت زمانبندیشده
✓ حفظ هویت کاراکتر و ثبات محیط در شاتها
✓ کنترل حرکت گلولهها و افکت توقف
✓ رفع خطاهای رایج و بهینهسازی خروجی
@ai_farshad
Please open Telegram to view this post
VIEW IN TELEGRAM
❤10👎2😁2👍1
آموزش هوش مصنوعی | ابزار و پرامپت تولید محتوا
راهنمای_ساخت_ویدیوی_توقف_گلوله_Seedance_2_5_1.pdf
555.2 KB
❤4👍2😁1
• ویدیوی اصلی را برای ادیت آماده کنید
• از چهار تصویر بهعنوان رفرنس استفاده کنید
• هر ویدیو را به شاتهای حداکثر ۱۰ ثانیه تقسیم کنید
• یک پرامپت مادر و قابلاستفاده برای پروژههای مختلف بسازید
• چهره، لباس، قاب و فضای هر پروژه را تغییر دهید
• هویت سوژه و حرکت طبیعی ویدیو را بهتر حفظ کنید
• خطاهای رایج Google Omni را شناسایی و اصلاح کنید
• خروجی شاتها را برای تدوین نهایی آماده کنید
📘 فرمت محصول: فایل PDF جامع ۳۸ صفحهای
https://pelymedia.com/product/google-omni-video-editing-tutorial/
همراه ما باشید. 🙏🏻
@ai_farshad
Please open Telegram to view this post
VIEW IN TELEGRAM
👍6❤5😁5
Create a hyper-realistic black and white fashion portrait based on the uploaded reference image. Keep the exact same composition, vertical framing, close portrait crop, camera angle, subject distance, background wall texture, shadow placement, lighting direction, contrast, film grain and minimal editorial atmosphere.
Replace only the original person with the character from the newly uploaded target reference image. Preserve the target character’s facial identity, face shape, skin tone values in monochrome, hairstyle, hair texture, eyebrows, eyes, nose bridge, visible upper face details, jawline if slightly visible, and natural expression as accurately as possible.
The character must keep the same pose: centered close portrait, head facing forward, eyes looking directly into the camera, lower half of the face covered by a thick black high collar, scarf or oversized knit fabric. Only the eyes, forehead, hair and upper nose area should be clearly visible. The expression must feel intense, calm, mysterious and editorial.
Maintain the exact same lighting: hard directional black and white studio light coming from the upper side, strong triangular shadow across the face and background, bright highlight on the eyes and upper face, deep black shadows on the lower face and clothing, high contrast monochrome with visible film grain.
Adapt the outfit to the target character while keeping the same oversized black high-neck knitwear / scarf / coat texture. The fabric should look soft, thick, matte, dark and premium. The collar must cover the mouth and lower face exactly like the reference.
Use a 50mm to 85mm portrait lens, close-up fashion photography, shallow depth of field, textured monochrome film look, strong shadow geometry, clean minimal background. No text, no logos, no extra objects, no smile, no color, no watermark.همراه ما باشید. 🙏🏻
@ai_farshad
Please open Telegram to view this post
VIEW IN TELEGRAM
❤14👍3😁3👎1
Use my uploaded photo as the exact identity reference. Preserve my real facial identity, facial structure, eyes, nose, lips, jawline, skin tone and natural facial features. Do not change my identity.
Create a photorealistic minimalist fashion portrait inspired by the reference image.
A young woman standing indoors against a plain warm light-gray/beige textured wall. She is wearing a voluminous oversized chunky teddy-textured knit sweater/cardigan in rich warm caramel brown, with an extremely thick fluffy boucle texture. The sweater wraps around the upper body and arms, creating a cozy oversized silhouette. Underneath, a simple fitted black high-neck top is subtly visible around the neckline.
She is also wearing a black beret tilted naturally on her head. Long straight dark-brown hair falls naturally from underneath the beret, framing her face and extending down her back.
Pose: her body is turned slightly toward the side, approximately three-quarter profile. Her head is turned toward the right side of the frame, looking off-camera with a calm, slightly thoughtful and elegant expression. Her arms are crossed naturally in front of her torso, partially hidden inside the oversized sweater sleeves. Relaxed shoulders, natural posture, effortless fashion pose.
Soft diffused indoor lighting with a subtle warm tone. Gentle light falls evenly across the face and sweater, with very soft shadows. Minimal background with no furniture or distracting objects. Slightly imperfect textured plaster wall.
Photography style: minimalist editorial fashion photography, cozy winter aesthetic, understated European street-style feeling, natural skin texture, realistic fabric fibers, detailed chunky knit texture, soft muted colors, subtle film grain, gentle contrast, authentic photographic look.
Vertical portrait composition, approximately 4:5 aspect ratio, medium shot from around the waist/chest upward. Subject positioned slightly off-center. 50mm lens look, natural perspective, subtle background softness.
Highly photorealistic, realistic anatomy, realistic hands, natural facial proportions, realistic hair strands and extremely detailed sweater fibers, candid but carefully composed fashion photograph.همراه ما باشید. 🙏🏻
@ai_farshad
Please open Telegram to view this post
VIEW IN TELEGRAM
❤23👍6😁1
@ai_farshad
Please open Telegram to view this post
VIEW IN TELEGRAM
❤8👍8
Use a imagem de referência como base e crie um retrato editorial ultra-realista do mesmo homem, no mesmo estilo e identidade.
Fotografia em close-up com ângulo de câmera oblíquo e diagonal, levemente acima e de um lado do rosto, como na referência.
O olhar deve ser direto para a câmera, com expressão séria, confiante e composta.
Vista um blazer preto sob uma camisa branca.
Iluminação de estúdio suave e controlada, com contraste cinematográfico, um lado do rosto mais iluminado e o outro em sombra suave.
Fundo minimalista e neutro.
Foto com lente de 85mm, alta definição, textura real da pele, detalhes de cabelo e barba, estilo editorial premium.
Negative prompt:
outra pessoa, rosto diferente, mudar cabelo, alterar identidade, cabelo raspado, pele artificial, filtro de beleza, plasticidade, distorção, cartoon, CGI, texto, logo, watermark.همراه ما باشید. 🙏🏻
@ai_farshad
Please open Telegram to view this post
VIEW IN TELEGRAM
❤22👍7😁3
🤖 OpenAI از GPT-6 Sol و Luna رونمایی کرد؛ خطای کمتر با نصف قیمت!
⬅️ تنها چند دقیقه بعد از معرفی Claude Opus 5.5، شرکت OpenAI از دو مدل جدید خانواده GPT-6 پرده برداشت.
✔️ GPT-6 Sol
✔️ GPT-6 Luna
📉 خطای کمتر، هزینه پایینتر
💰 قیمتها هم تا ۵۰٪ کاهش پیدا کرده:
🔺 Sol
🔺 Luna
♻️ مدلهای جدید در ChatGPT Work، Codex و API در دسترس قرار گرفتهاند.
📌 به نظر میرسه رقابت مدلهای هوش مصنوعی وارد مرحله جدیدی شده؛ قدرت بیشتر، هزینه کمتر و سرعت بالاتر.
@ai_farshad
برای کارهای پیچیده و نیازمند استدلال پیشرفته طراحی شده؛ از جمله برنامهنویسی، تحلیل و پروژههای حرفهای.
تمرکز این مدل روی کارهای سریع، حجیم و تکراریه؛ مثل خلاصهسازی اسناد، استخراج اطلاعات و پاسخ به پرسشهای سادهتر.
طبق ارزیابیهای داخلی OpenAI، میزان خطای Sol نسبت به نسل قبلی تقریباً نصف شده و عملکرد آن از نظر قابلیت اطمینان به GPT-6 Astra نزدیک شده است.
💰 قیمتها هم تا ۵۰٪ کاهش پیدا کرده:
• ورودی: ۲ دلار / ۱ میلیون توکن
• خروجی: ۱۰ دلار / ۱ میلیون توکن
• ورودی: ۰٫۱ دلار / ۱ میلیون توکن
• خروجی: ۰٫۵ دلار / ۱ میلیون توکن
♻️ مدلهای جدید در ChatGPT Work، Codex و API در دسترس قرار گرفتهاند.
همچنین Luna برای کاربران دسکتاپ و کاربران پلنهای Free و Go عرضه خواهد شد.
@ai_farshad
Please open Telegram to view this post
VIEW IN TELEGRAM
❤9👍9
Use the attached face image as the exact identity reference and generate the same person.
IDENTITY LOCK — highest priority:
Preserve the exact identity, face shape, facial proportions, eyes, eyebrows, nose, lips, jaw, ears, skin tone, facial hair, tattoos and natural imperfections from the uploaded reference.
HAIR LOCK:
Do not modify the natural hair in any way. Preserve identical hairline, shape, volume, texture, length and density.
Create an ultra-realistic close portrait of the subject wearing a loose off-white linen shirt in front of a warm neutral beige background.
Frame from the mid-chest upward.
The subject is slightly angled toward the camera. One forearm naturally crosses the lower part of the frame, with the hand loosely resting near the opposite upper arm.
The gesture must look casual and spontaneous, not like a fashion pose.
His head is slightly tilted and his eyes look directly toward the camera.
Expression is neutral, serious, calm and masculine, with relaxed lips and no smile.
Use soft natural side window light, revealing pores, beard texture, tattoos, natural skin shine and subtle facial imperfections.
The image should feel intimate and realistic, like a candid portrait captured indoors during daylight.
No elaborate styling, no props, no dramatic background.
Negative prompt:
exaggerated pose, crossed-arm business pose, smiling, feminine hand pose, beauty filter, retouched skin, glossy artificial face, changed identity, changed hairstyle, CGIهمراه ما باشید. 🙏🏻
@ai_farshad
Please open Telegram to view this post
VIEW IN TELEGRAM
❤17👍5😁2
Use the attached face image as the exact identity reference and generate the same person.
IDENTITY LOCK — highest priority:
Preserve the exact identity, face shape, facial proportions, eyes, eyebrows, nose, lips, jaw, ears, skin tone, facial hair, tattoos and natural imperfections from the uploaded reference.
HAIR LOCK:
Do not modify the natural hair in any way. Preserve identical hairline, shape, volume, texture, length and density.
Create an ultra-realistic close portrait of the subject wearing a loose off-white linen shirt in front of a warm neutral beige background.
Frame from the mid-chest upward.
The subject is slightly angled toward the camera. One forearm naturally crosses the lower part of the frame, with the hand loosely resting near the opposite upper arm.
The gesture must look casual and spontaneous, not like a fashion pose.
His head is slightly tilted and his eyes look directly toward the camera.
Expression is neutral, serious, calm and masculine, with relaxed lips and no smile.
Use soft natural side window light, revealing pores, beard texture, tattoos, natural skin shine and subtle facial imperfections.
The image should feel intimate and realistic, like a candid portrait captured indoors during daylight.
No elaborate styling, no props, no dramatic background.
Negative prompt:
exaggerated pose, crossed-arm business pose, smiling, feminine hand pose, beauty filter, retouched skin, glossy artificial face, changed identity, changed hairstyle, CGIهمراه ما باشید. 🙏🏻
@ai_farshad
Please open Telegram to view this post
VIEW IN TELEGRAM
❤14👎9😁6👍4