Large-Scale Image Dataset Collection for AI Training

We’re looking for responsible and creative individuals to participate in a high-quality image data collection project for training next-generation AI systems.

This project focuses on real, physical photography — no AI generation, no Photoshop, no digital manipulation. If you enjoy structured shoots, attention to detail, and creating clean, consistent visual data, this project is for you.

🔍 Project Overview

The project involves capturing paired and sequential images across 5 distinct categories, each designed to represent realistic visual changes, object relationships, and human activities.

Applicants are expected to deliver a minimum of 250 photos per production round across all five categories.
Each category has clearly defined technical and creative requirements that must be followed precisely to ensure consistency and quality across the dataset.

📦 Deliverables: High-resolution JPEG images
🎯 Purpose: Training and evaluation of AI visual understanding models
🌍 Locations: Real indoor and outdoor environments
🛠 Editing: Physical changes only – absolutely no digital edits

🚫 Critical Global Rules (Read First)

Failure to follow these rules will result in immediate rejection:

NO AI & NO PHOTOSHOP
All changes must be done physically. No filters, retouching, object removal, or generation.
TRIPOD REQUIRED (Categories A, C, D)
Camera must remain completely stationary.
Use a timer or remote shutter – no hand-pressing.
NO BRANDS OR IP
No visible logos, brand names, license plates, or recognizable trademarks.
METADATA IN ENGLISH ONLY
File names and descriptions must follow the provided naming convention.

⚙️ Technical Requirements (All Categories)

Format: JPEG, sRGB
File size: 4 MB – 50 MB
Aspect ratio: 4:3, 16:9, or 1:1 (consistent within each series)
File naming: sceneXX_shotXX_description.jpg

✅ Who This Is Perfect For

People with a genuine passion for photography or a strong visual instinct for what makes a high-quality, well-composed image (lighting, framing, depth, and clarity all matter).
Detail-oriented contributors who can deliver consistent results across multiple images, ensuring accuracy in framing, lighting, background alignment, and metadata.
Reliable and organized individuals who can manage file creating and naming, shot sequencing, and documentation without missing steps.
Professionals or advanced hobbyists with access to a stable shooting setup (tripod, controlled environments, and basic lighting awareness).