AI Synthetic Dataset Generation
Budget: $30 – $250 USD
I'm building an AI product that requires a synthetic dataset for inpainting tasks. The goal is to generate images of furnished rooms, corresponding object masks, and clean empty-room targets.
Dataset Structure:
```
dataset/
image/ # input room images (with furniture)
mask/ # masks of removable objects (sofa, bed, chairs, tables, etc.)
target/ # ground truth empty room images (without furniture)
```
Requirements:
- Around 5000 samples.
- Each sample must include:
- Input image (room with furniture).
- Object mask(s) (sofa, bed, chair, table, lamp, TV, wardrobe, etc.).
- Ground truth empty room (same room but without furniture).
- Dataset must be consistent and high quality (realistic rooms, masks well aligned).
- Ideally generated using Stable Diffusion / ControlNet / LoRA or other reliable methods.
- Final delivery: structured dataset in folders + metadata (JSON/CSV).
- Image format: PNG
Ideal Skills and Experience:
- Proficiency in Stable Diffusion or similar tools.
- Experience in generating synthetic datasets.
- Attention to detail for high-quality, well-aligned masks.
Dataset Structure:
```
dataset/
image/ # input room images (with furniture)
mask/ # masks of removable objects (sofa, bed, chairs, tables, etc.)
target/ # ground truth empty room images (without furniture)
```
Requirements:
- Around 5000 samples.
- Each sample must include:
- Input image (room with furniture).
- Object mask(s) (sofa, bed, chair, table, lamp, TV, wardrobe, etc.).
- Ground truth empty room (same room but without furniture).
- Dataset must be consistent and high quality (realistic rooms, masks well aligned).
- Ideally generated using Stable Diffusion / ControlNet / LoRA or other reliable methods.
- Final delivery: structured dataset in folders + metadata (JSON/CSV).
- Image format: PNG
Ideal Skills and Experience:
- Proficiency in Stable Diffusion or similar tools.
- Experience in generating synthetic datasets.
- Attention to detail for high-quality, well-aligned masks.