Fetch dollhouse
View all dollhouse nodes from Neo4j. Click one to see JSON, render 3D model preview, and view images.
Dollhouse list
No dollhouses yet
Enter your email and click List to load your dollhousesDollhouse Automation
Upload files and run all 4 steps in one click: Create → Extract Objects → Convert to 3D → Analyse
⚡Dollhouse Automation
Login first to auto-fill, or paste token manually
Render dollhouse
Enter a dollhouse ID to render captured images and view the model path.
Render Configuration
My Projects
Create and manage interior design projects. Link DPP profiles, dollhouses, and build shopping lists with cost estimates.
Your Projects
No projects yet
Enter your email below and click Load My ProjectsAPI Explorer
List of all APIs used in this repo. Click Test to send a request with your payload and view the response.
Interior Analysis
Upload a room or interior image. Click Analyze to get scene (style, palette, lighting) and a list of objects with dimensions, materials, colors, fabric, texture, and finish — for interior designers.
1Select image
Click or drag an image here
PNG · JPG · WebP — max 5 MBObject Placement
Upload a room image + one or more object images + a placement instruction. AI places the objects into the room photorealistically using Nano Banana — gemini-2.5-flash-image via ProcessIQ.
1Room image
Click or drag the room image here
PNG · JPG · WebP — max 5 MB2Object images (one or more)
Click or drag object images here
Add multiple objects — each max 5 MB3Placement instruction
Shapely Layout Validation
Upload a JSON file from OpenCV object detection (room boundary + objects like sofa, chair, coffee_table with coordinates). The backend uses Shapely to validate the layout against interior design rules (min clearance, no overlap, coffee table–sofa distance) and returns human-readable explanations plus raw violations.
1 Upload layout JSON
Click or drag a JSON file here
OpenCV-style layout: room + objects with coords (e.g. sofa, chair, coffee_table)Data Ingestion
Choose a source: Amazon, Google Images, Wayfair, or add a local vendor product manually. All results are saved to your Neo4j catalog.
Select Source
Catalog & 3D Conversion
Browse all items in Neo4j. Convert pending images to 3D models via Meshy AI.
No items yet
Fetch images to populate the catalog3D Model Preview
Search Product
Search the GaaZoo catalog. If no results are found, items are automatically fetched from Amazon and Google Images and saved.
Search filters
Smart Search (AI-Powered)
AI-powered product search using Template 37. Provide intent, budget, DPP, dollhouse IDs, and notes. AI analyzes all inputs and returns product recommendations with intent classification (replace/add/remove/redesign).
Search Parameters
2D → 3D model converter
Upload any product image and generate a 3D GLB model directly via Meshy AI. No catalog required.
1Select image
Click or drag an image here
PNG · JPG · WebP — max 10 MB
2
Real‑world dimensions
(optional)
Width (X) · Height (Y) · Depth (Z). Leave blank to keep Meshy's default scale. Partial values are accepted — unspecified axes are scaled proportionally.
3D model ready
Resize and Scale 3D model
Upload a .glb or .obj file and optionally scale (one dimension) or resize (two or three dimensions) to real-world size. No Meshy API needed.
1Select 3D file
Click or drag a .glb or .obj file here
GLB · OBJ — max 50 MB
2
Scale or resize
(optional)
Width (X) · Height (Y) · Depth (Z). Leave blank to view at original scale. One value = scale (proportional, no skew). Two or three = resize only those dimensions; others stay at original size.
3
Scale by %
(optional)
Scale the model uniformly by a percentage. Increase = enlarge; decrease = shrink.
Model loaded