📸 Facial Image Collection Project

Share Your Facial Images. Help Build Inclusive AI. Earn Rewards for Approved Submissions.

💡 Project Overview

We are launching a large-scale biometric facial image collection project to build a diverse dataset across multiple countries, ethnicities, age groups, and genders.

Freelancers can submit one or both datasets for each unique participant. Partial submissions are fully welcome.

• Set A Only – Age Progression Dataset (15 Images)

• Set B Only – Facial Mask Dataset (20 Images)

• Both Sets – Age Progression + Facial Mask (35 Images Total)

⚠️ Important: All images must be original, high quality, and belong to a single, unique participant. Submissions must successfully pass our Quality Assurance (QA) approval workflow.

🌍 Participant Eligibility & Age Rules

• Minimum Age: All participants must be 18 years or older.

• Compliance: Explicit participant consent is mandatory before submitting any data.

Set A (Age Progression) Strict Age Brakets:

• Aged 23+: Fully eligible.

• Aged 21–22: Eligible only if all historical photos were taken when the participant was 18 years or older.

• Aged 18–20: Not eligible for Set A (but fully welcome to participate in Set B).

🕒 Set A: Age Progression Dataset (15 Images)

This dataset consists of 15 historical photos of the same individual taken across different points in time.

• Accepted Sources: Phone gallery, Google Photos/iCloud, WhatsApp backups, personal social media, or digitized printed photographs.

✅ Requirements

• All 15 photos must belong to the exact same participant.

• Minimum 1-month gap between any two photos.

• Maximum 5-year span between the oldest and newest photo.

• Face must be frontal, centered, and clearly visible.

• A diverse mix of locations, occasions, and lighting (indoor/outdoor) is preferred.

• Maximum of one sunglasses image allowed.

❌ Strict Rejections (Not Allowed)

• Selfies, edited images, or beauty mode filters.

• AI-generated images, stock photos, or downloaded celebrity images.

• Photos of printed photographs captured using another phone camera.

• Duplicate photos or repetitive backgrounds.

😷 Set B: Facial Mask Dataset (20 Images)

This dataset requires 10 images in a Surgical Mask and 10 images in a Cloth Mask (masks can be reused across shots).

✅ Requirements

• Mask must fully cover both the nose and mouth.

• Face must be frontal, centered, and sharply in focus.

• Incorporate varied head positions with slight angles (Yaw and Tilt).

• Use different locations and backgrounds across all 20 images.

• Include a mix of indoor and outdoor lighting.

• Maximum of one sunglasses image allowed.

❌ Strict Rejections (Not Allowed)

• Selfies, screenshots, or internet/AI-generated images.

• Filters, beauty mode, or post-capture edits.

• Blurry, low-quality, or poorly lit photos.

• Visible background clutter like other people, pets, ID cards, or license plates.

• Repeated or highly similar backgrounds.

📷 Technical & Quality Specifications

• Format: Original JPG or PNG files only.

• Minimum Resolution: 320 × 240 pixels.

• Minimum Face Height: Face must span at least 240 pixels vertically within the frame.

• Framing: Face must be completely visible and not cropped at the edges.

❌ Common Rejection Summary: Selfies or incorrect face positioning; side profiles or tilted faces; motion blur; text overlays, date stamps, or watermarks; identity mismatches across a single set; or masks slipping below the nose.

📂 Required Google Drive Folder Structure

Do not upload files directly to the submission form. Organize your files in Google Drive exactly as shown below:

If Submitting Set B (Facial Mask):

Plaintext

[Participant Name]

└── Facial Mask Set

    ├── Surgical Mask Set (10 Images)

    └── Cloth Mask Set (10 Images)

If Submitting Set A (Age Progression):

Plaintext

[Participant Name]

└── Age Progression Image Set (15 Images)

🔒 Crucial Link Setting: Before submitting, ensure the Google Drive folder link permission is explicitly set to "Anyone with the link can view".

Important Links