E-commerce Garment Image Pipeline Development

What we're building A pipeline that takes one garment design and one model photo, composites them once into a clean "model wearing garment" master shot, then generates the 7 standard e-commerce angles above from that same composite not from scratch each time so the model's identity and the garment stay consistent across the whole set, ready to assemble into a lookbook spread. Problem it solves Fashion e-commerce needs the same product shown from 6-7 angles per listing, which normally means a full reshoot per garment. Regenerating each angle independently with AI creates a worse problem a different-looking "model" in every shot. This workflow fixes that by locking the angle-generation step to edit one fixed reference image rather than re-imagining the scene each time.

Workflow Steps Step 1 - Garment Design (Image Node) Upload the garment reference flat lay, mannequin shot, or sketch. Label "Garment Design." Step 2 - Model Reference (Image Node) Upload the model photo neutral pose, plain background. Label "Model Reference." Step 3 - Composite Request (Text-to-Text Node, Claude Opus 5) Connect both Garment Design (Step 1) and Model Reference (Step 2). Prompt: "Change the dress of the model in image 1 from the current outfit to the new dress design in image 2, ensuring the new dress style is reflective of the provided dress details." Step 4 - Generate the Composite (Image Node, Nano Banana 2) Connected to Step 3's output. Produces "Outfit Swap Overview" the model now wearing the garment, full body, front-facing. This becomes the one fixed reference every angle below derives from. Step 5 - Angle Brief Writing, per angle (Text-to-Text Node, Claude Sonnet 5) One text node per angle, each connected to the Outfit Swap Overview (Step 4). For each, Claude is prompted with a short angle instruction Step 6 - Render each angle (Image Node, Nano Banana 2) Each connected to its matching Step 5 brief and the Outfit Swap Overview (Step 4). Produces: "3/4 Angle Variation," "Back View Display" → back render, "Garment Silhouette View" → side profile, "Textile Details" → fabric close-up, "Fashion Motion Shot" → walking pose, "Model Close-Up" → neckline/collar detail. Step 8 - Quality check across the set Scan all 6 outputs for identity drift (same model recognizable) and garment consistency (same construction, fabric, color across every angle). Step 10 - Export

Workflow Project link - https://app.flora.ai/join-project/e75d89bc-69b4-42f2-ab2b-7805ba93909c