Visual Factory provides a one-stop service for creating e-commerce mockups and videos using AI.

In today's digital age, competition in the e-commerce industry is becoming increasingly fierce. How to...High efficiencyFurthermore, creating engaging clothing display content at low cost has become a key need for businesses. This article will explore an innovative solution—using...Visual FactoryAItoolChoose Banana AIGenerate outfit models via Veo AITransform it into a high-quality, dynamic presentation video. Combined with...AITechnology can bring significant advantages to apparel e-commerce by reducing costs and increasing efficiency, and can create more visually impactful and attractive viral content. Let's learn how to achieve this amazing creative process step by step.

Visit the Visual Factory website https://d.aijuh.com/, register an account and log in.

Depending on your needs, you can choose Banana's graphing features.

allPrompt wordsUse it directly in the banner image editor. Upload the original clothing image (or a photo of a fitting model), then copy the following...Prompt wordsPaste the image and select 9:16 to generate continuous images.

femaleAIModel changing clothes

Prompt wordsE-commerce product display images, 8K ultra-high definition, commercial photography quality, soft and high-end indoor lighting. Main elements: 1. Live model: A young East Asian woman (long, slightly wavy hair, natural light makeup), wearing [uploaded image 1] (reference image weight 0.95, feature locked + detail threshold 98, artistic correction 0.1); pose: sideways, lightly touching her hair, holding a phone for a selfie, relaxed and natural. 2. Human-shaped mannequin: A 1:1 scale **complete white 3D mannequin (with head + complete limbs + lower body)**, wearing the exact same clothing as the live model, standing on a **metal stand with a base** (the stand only exposes the legs, the lower body is covered by the mannequin), located to the right of the live model, in a standard standing pose, with clothing style/details completely identical to the live model. Background Scene: A minimalist clothing studio backdrop, with a **fashion show runway image** projected onto the wall. The model in the image is the same person as the main live model and wears the exact same clothing (the runway is light gray, and the audience is blurred). A black clothes rack (holding clothing in the same color family), a stone-textured tabletop, and an Apple computer are visible on the left. A black beanbag chair is on the right, and the floor is light gray tile. The space is clean and bright. Mandatory Requirements: – The mannequin must be a complete human form (including head, limbs, and lower body), and must not be headless/only upper body/exposed metal rods; – The runway model in the background projection is the main live model plus the same clothing, with completely identical appearance and attire; – The live model faces the camera directly throughout, without any sideways or tilted angles. Selfies taken with a handheld phone that partially obscure the face have a natural effect; – Clothing details (texture, pattern, accessories) are 100% reproduced, and the clothing worn by the live model is completely identical to that of the model; – The focus of the image is on the clothing, with a slightly blurred background, conforming to the sophisticated aesthetics of e-commerce product images.

Prompt wordsE-commerce product display images, 8K ultra-high definition, commercial photography quality, soft, high-end custom lighting (matching warm and cool tones to the clothing style). Prioritized in-depth analysis of the core information in [Uploaded Image 1]:automaticIdentify clothing styles (retro/street/commuter/sweet and cool/affordable luxury, etc.), cut designs, color schemes, material textures, and core elements. Based on this, match highly compatible scenes and visual tones. Reference image weight: 0.95; feature locking + detail threshold: 98; artistic correction: 0.1. Main elements: Real-life model: Young East Asian woman (long, slightly wavy hair, natural light makeup), wearing the clothing in [uploaded image 1], incorporating a subtle pheromone effect—soft, expressive eyes, slightly pursed lips with a smile, facing the camera with her hair lightly touching it (relaxed and unforced hand movements), holding her phone to naturally partially obscure her face. Her overall temperament is a balance of sweetness and allure, restrained yet attractive, with a relaxed and natural demeanor. Humanoid Mannequin: A 1:1 scale, complete white three-dimensional mannequin (including head, complete limbs, and lower body), dressed in clothing exactly the same as the real person, standing on a metal stand with a base (the stand only exposes the legs, the lower body is covered by the mannequin), located to the right of the real person, in a standard standing posture, with 100% consistency with the real person's clothing pattern/details. Background Scene: Dynamically matches the scene based on the clothing style (replacing a single studio): For retro style: Use an old Shanghai alleyway/retro cafe background, projecting a Republic-era runway show of the same clothing worn by the models onto the wall; For street style: Use an urban trendy street/industrial warehouse background, projecting a street fashion show of the same clothing worn by the models onto the wall; For commuter style: Use a light luxury office building/high-end meeting room background, projecting a business fashion week runway show of the same clothing worn by the models onto the wall; For sweet and cool style: Use a trendy toy store/neon street corner background, projecting a sweet and cool style runway show of the same clothing worn by the models onto the wall; General Details: In the projected runway show, the models are the main live models, and the clothing is the same (the runway color scheme matches the scene style, and the audience is blurred); Display props matching the clothing style are placed on the left (retro style: wooden hangers + old-fashioned gramophone / street style: metal hangers + trendy skateboard / commuter style: leather hangers +...). Luxury decorative items), stone-patterned countertop + Apple computer; soft furnishings matching the scene are placed on the right (retro style with velvet sofa / street style with industrial style wrought iron chair / commuter style with genuine leather single sofa); the floor material matches the scene tone (retro style with solid wood floor / street style with cement floor / commuter style with marble floor), the space is clean and transparent, and the background is slightly blurred to highlight the clothing. Mandatory Requirements: Human figures must be in "complete human form (including head, limbs, and lower body)," and must not be "headless/only upper body/exposed metal rods"; The runway model in the background projection is the main live model plus the same clothing, with completely identical appearance and attire; The live model faces the camera directly throughout, without any sideways or tilted angles, and holding a phone to cover the face creates a natural effect; Clothing details (texture, pattern, accessories) are 100% reproduced, and the clothing on the live model is completely identical; The focus of the image is on the clothing, with a slightly blurred background, conforming to the sophisticated aesthetics of e-commerce product images; The overall visual style is highly consistent with the style of the clothing in [Uploaded Image 1], with the scene serving the style of the clothing and not deviating from the characteristics of the clothing itself.

maleAIModel changing clothes

Banana generates outfits.AIThe model can be a real person's portrait, such as one's own employees or model images, divided into single finished images or multiple unfinished images, and then combined into a new one.AILive models.

Hyper-realistic e-commerce clothing product display, forcibly locking the clothing features in [uploaded image 1] (reference image weight 1.0, feature locking threshold 100%, artistic correction 0, prohibited).AI(Auto-generated clothing), the main character and the humanoid mannequin beside him **100% replicate the style, texture, pattern, and accessories of the clothing in the picture**, without any deviation; Main character setting: a young East Asian male, **fully presented in a 9:16 aspect ratio**, facing the camera directly the entire time, without any side or tilt angles, with a clearly visible face and a natural and friendly , and a generous and relaxed posture (randomly generated: ① gently adjusting the collar/cuffs of the same clothing as the humanoid mannequin with both hands; ② standing naturally with the elbow lightly resting on the shoulder of the humanoid mannequin; ③ one hand supporting the waistline of the humanoid mannequin, with the other hand hanging naturally); Humanoid mannequin requirements: **complete and intact, with a head model, proportionate**, completely identical to the main character's clothing, placed to the right of the main character, **placed on a small display stand/staggered base to create a natural height difference with the main character, visually focusing on the main character**; Scene and Lighting: The scene strictly matches the clothing style (casual casual style → minimalist trendy store/commuter style → light luxury business store/winter style → retro warm style store/sports style → sports brand store/resort style → island style boutique). **The scene constructs a sense of spatial depth (using staggered shelf heights, perspective relationships between doors and windows, and blurred foreground props to enhance layering). The background uses a gradual blurring (clear foreground, soft midground, and hazy background)** to highlight the main character and clothing. Commercial lighting is adapted to the texture of the clothing (bright natural light for casual style/cool sophisticated light for professional style/warm yellow soft light for winter style). The main light source focuses on the main subject, and the auxiliary light weakens the background shadows to enhance the immersive feeling of the image. Image Quality Requirements: 8K high definition, natural and sophisticated colors, in line with the high-end aesthetic of e-commerce product images, and the overall visual style is highly consistent with [uploaded image 1]. Mandatory Constraints: No irrelevant elements are allowed (such as abrupt props, extra characters, cluttered background decorations, etc.). Only clothing, complete humanoid mannequin props, and core elements matching the scene are retained.

Disassembling clothes

Prompt wordsDisassemble the character's clothes.

Realize e-commerceAIThe core technique of marketing videos: the "core dynamic + static extension" material combination method, such as how to plan and generate two sets of materials to achieve zero editing.fastComplete film

• TargetBreaking the 8-second limit for a single video on Veo.High efficiencyGenerate a set of "semi-finished" video footage that can be directly spliced together without frame-by-frame editing.
• Core ideaWe do not pursue using a longPrompt wordsInstead of generating complex videos, it's about designing "shots" like a director, allowing...AIGenerate them separately, and then use them in the editing software.Simplesplicing.
• Two sets of design materials:
  • Core dynamic shots (0-8 seconds): Used for opening or transitions to attract attention and showcase dynamic try-on effects.
  • Static extended shots (each 0-8 seconds): Used to showcase the main subject, generating multiple clips from different angles or perspectives, serving as the main content of the video.

  Planning tips: Ensure that the model's pose, expression, and lighting are highly consistent in all "static extension" shots so that the stitching will not look abrupt.

  
  femaleAIModel Videos

  Option A: Generate "Core Dynamic Shot" + Option B: Generate "Static Extended Shot"

  veoPrompt words- A: Core Motion Lens (General Version)

  The woman holds a phone, completely obscuring her face; her eyes shouldn't be too frightening, but the occasional glimpses of her eyes should be attractive. She moves slowly, sideways, forward, and backward, her other hand naturally smoothing her hair, making a peace sign, and adjusting her clothes. Details throughout naturally showcase the cut and texture of her outfit; her movements are unhurried and fluid, her posture relaxed, elegant, sweet, and charming. Her shadow should follow her, and the image should be realistic and high-definition.

  veoPrompt words-B: Still image extension (freeze-frame photograph)

  Create an 8-second 8K ultra-high-definition fashion video, generated by Veo3, with no audio and no watermark. –style "AI Matching the style of uploaded materials, pure static snapshot photography quality" –character "Refer to the image of uploaded materials, enhance sweetness and cuteness (round eyes / energetic smile / soft facial features)" –motion "0-8 seconds of pure static shutter freeze (no limbs / props / slight scene movement)" –detail "Ultra-fine texture, clear panoramic view, no hair / cloth movement in the freeze frame, no outlines / special effects" –colorAI Matching the color scheme of the material, soft lighting to enhance the face –duration 8.0 –transitions 0-8 seconds 8 sets of panoramic freeze frames (1 second each), 0.1 second random fast transition (left shift/left squeeze) –shot-type Full panoramic full body, close-ups disabled –shutter-effect Shutter capture static freeze frames –static-poses-scene 8 sets of freeze poses + scene adaptation: frontal natural standing pose (with a simple solid color background wall, highlighting the overall outline of the clothing, suitable for minimalist/professional style, centered panoramic display of the person); 45° side standing pose (with a stylized floor-standing clothes rack, with accessories of the same style hanging on the rack, suitable for casual/retro style, highlighting the side cut of the clothing); 90° side standing pose (with a minimalist side table/ornament, with stylized small objects on the side table: warm style with knitted cushions, retro style with brass candlesticks, suitable for warm/industrial style, showcasing the drape of the clothing); Slightly turned to the side (paired with a key prop, such as a coffee cup/briefcase on a table, the props and style echo each other: a coffee cup for a casual style, a briefcase for a professional style, enhancing the sense of immersion in the scene); 45° frontal standing posture (paired with a blank background in the corner of the scene, the blank background highlights the character, suitable for street photography/casual style, creating a sense of space); Frontal standing posture (paired in front of the main visual background, such as a bay window/floor-to-ceiling window, the bay window is covered with a blanket of the same color (warm style), the floor-to-ceiling window is placed next to a simple office chair (professional style), suitable for warm/professional styles, natural light outlines the texture of the clothing); Half-turned leaning posture (paired next to a retro sofa/wrought iron shelf, the body half-leans against the static furniture, one hand lightly rests on the backrest/edge of the shelf, retro books are decorated next to the sofa/green plants on the wrought iron shelf, suitable for retro/industrial style); Frontal standing posture with arms crossed (paired with an industrial style metal door frame/home entryway background, the arms are naturally crossed at the abdomen, suitable for professional/ Minimalist style, highlighting the clean lines/gentle texture of the clothing. –props-scene "Props/scenes are static throughout, placed according to the character's pose but without any animation, not obscuring the character/clothing, scene details match the clothing style." –clothing-auto-optimize "Restores the full view of the uploaded clothing." –no-effects "Removes outlines/highlights/unnecessary effects."

  maleAIModel Videos

  Option A: Generate "Core Dynamic Shot" + "Outdoor Scene" + Option B: Generate "Static Extended Shot"

  veoPrompt words- A: Core Motion Lens (General Version)

  Based on the uploaded reference image,automaticIt identifies clothing pattern/fabric/style, restores the main body 1:1, and automatically matches and adapts accessories, handheld objects and immersive scenes to generate an 8-second high-quality outfit display video; the storyboard is designed around the details of the clothing, the camera movement is natural, the lighting and shadow match the tone of the reference image, with a movie-level texture, 24fps, 1080p, and with soft ambient sound effects.

  veoPrompt words- A: Outdoor Scenarios (General Version)

  Based on the uploaded reference image, the clothing of Asian men in their 20s is accurately reproduced in a 1:1 ratio, including the cut, fabric texture, color saturation, and details.AI Automatically match accessories, bags, and handheld items to create an 8-second high-end outfit showcase video.
  Core Requirements
  Random combination:AI Choose from various scenes such as urban streets, artsy districts, and retro cafes. Randomly select the following shots in no fixed order, seamlessly connecting them within 8 seconds. Each generated shot and scene combination will be unique:
  Full-body dynamic shots: the subject walks/turns, with handheld tracking shots from the side and rear, showcasing clothing proportions, the swaying of trousers, and the arc of a bag's movement.
  Close-up of the upper body: Stand in front of a minimalist background, smile and look directly at the camera, or fidget with small objects, focusing on details such as the neckline, accessories, and fabric.
  Dynamic snapshots: a playful glance back at the camera as the character walks down stairs/slopes, and panning shots to capture the movement of clothing and bags.
  Light and shadow texture: Posing naturally under the dappled sunlight of trees or the interplay of light and shadow on buildings, shooting from a low angle to highlight the texture of fabrics and the details of accessories.
  Scene blending shots: Walking through stylized scenes, following the viewer in panoramic shots to showcase the outfit's suitability for the environment.
  The final shot freezes: the figure looks up, smiles, and stares directly at the camera; the camera slowly pulls back, capturing the entire pose in a panoramic shot.
  Prioritize details: Highlight the sheen of the clothing fabric, the drape of the trousers, and the details of the shoes; accessories should not overshadow the main elements; and the subtle expressions of the characters should be natural and vivid.
  Visual specifications: Cinematic quality, 1080p/24fps, accurate depth of field, lighting and shadows perfectly match the atmosphere of the reference image, accompanied by gentle ambient sound effects.
  Core instructions: Prioritize ensuring zero deviation in clothing details, focus on immersive outfit display, randomly combine shots and scenes, and submit multiple times to generate different versions of the video.

  veoPrompt words-B: Still image extension (freeze-frame photograph)
  Upload a reference image; the image will 100% reproduce the facial features, hairstyle, physique, and complete clothing of an Asian man in his 20s.automaticMatches style-appropriate items (1-2 random items from cameras, phones, bracelets, and collectibles); forces no dynamic elements (no character movement, no background movement, no change in lighting), generating 8 purely static fashion photos with an atmospheric feel. Employs a creative transition using **"close-up details of clothing" transitions** (each photo pauses for 1-2 seconds, transition is 0.2 seconds, total duration adapted to an 8-second video, 1 frame/second, 1080p): During transitions, the highlights of the clothing in the previous image are magnified (knit texture/trouser lines/accessory quality), then a quick cut to the next complete portrait is made, creating a fast-paced and eye-catching effect that emphasizes the clothing's selling points; each generation generates random shot order + variable options, ensuring multiple versions are unique; no text or sound effects, warm and soft natural lighting + low-saturation sophisticated color scheme highlight the clothing's texture and how it looks on the body, creating a high-definition, ultra-realistic magazine-quality feel that stimulates the desire to buy.
  Choose 1 from 2 shot variables:
  Standing sideways in front of a portrait poster, holding a coffee | Poster [Black and White / Composite], Light [Soft Left / Soft Right]
  Close-up of objects held under trees in autumn | Leaves [Ginkgo / Maple Leaves], objects [Mobile Phone / Camera]
  Smiling, holding a cup in front of a cement wall | The wall surface [lightly mottled / heavily weathered], the smile [a slight smile / a pursed-lip smile]
  Looking back at the stone steps | Stone steps [with moss/dry], angle [45°/60°]
  Park stone bench, chin resting on hand | Stone bench [granite / marble], light spots [circular / starburst]
  Walking with your back to the glass curtain wall | Reflecting [blue sky/high-rise buildings] and the ground [marble/asphalt]
  Close-up of face touching the cup | Lip appearance [moist/matte], background [cool gray/light blue]
  Shot from below under the tree canopy | Leaves [dense/sparse], eyes [slightly closed/closed]

  Looking at the final result, the scene transitions between the front and back streets quite naturally.

  The "static and dynamic" combination of Banana + Veo allows apparel e-commerce to completely say goodbye to high-priced models and cumbersome studio shooting: a single flat lay image →AI Model → 9:16 Dynamic blockbuster, done in ten minutes, cost less than a cup of milk tea.

  The tutorial provides 8K Prompt wordsThe weight parameters and scene formulas can be directly copied to produce commercial-grade quality, truly realizing "anyone can shoot, even with zero experience".

  Quickly open Banana, upload your first outfit, and use Veo to bring it to life. The next viral product-selling video could be yours!

  Don't forget to come back and hand in your homework, let's do it together. AI The style of wearing clothes has reached new heights.

  Visual Factory has morepracticalThe gameplay awaits you to unlock.

  Official websitehttps://d.aijuh.com/