Large-Scale Selfie & Historical Photo Collection
Budget / Salary$10–30
TypeFreelance project
LocationRemote
Posted2 hours ago
We are looking for remote contributors based in South Asia, Africa, East & Southeast Asia, Europe, South America, North America, Middle East, and Australia to participate in a large-scale facial data collection project to support next-generation machine learning and identity verification systems.
What You Will Do (Total of 24–30 Images):
You will need to capture and upload a structured set of photos using your smartphone:
Current Front-Facing Selfies (4–6 photos): Captured at arm's length under clear, bright lighting with a natural or slightly smiling expression.
Current Head-Pose Selfies (4–6 photos): Captures of your face rotated at specific angles (tilting up/down or turning left/right).
Historical Photos (12–18 photos): Genuine, older pictures of yourself from your personal gallery. These must span a timeline of at least 1 year up to 20 years to help train models on natural aging.
Crucial Quality Standards (Submissions failing these will be rejected):
No Filters or Edits: Do NOT use portrait mode, beauty filters, skin smoothing, or any photo editing software. All photos must be completely raw and natural.
Original Metadata (EXIF): Your selfie files must contain their original camera data (date, time, and device details). No screenshots or scans of printed photos are allowed.
Clear Face Visibility: No sunglasses, masks, hats, or hair blocking your eyes, nose, or mouth.
Technical Specs: Images must be in JPEG format, with a minimum resolution of 512x512 pixels for the face area, and file sizes between 128 KB and 300 MB.
Payout & Compensation:
High payout rates are offered for every fully validated and accepted image dataset (up to 24 billable images per participant). If you are interested, please apply with your location and smartphone model. Rate is negotiable.
What You Will Do (Total of 24–30 Images):
You will need to capture and upload a structured set of photos using your smartphone:
Current Front-Facing Selfies (4–6 photos): Captured at arm's length under clear, bright lighting with a natural or slightly smiling expression.
Current Head-Pose Selfies (4–6 photos): Captures of your face rotated at specific angles (tilting up/down or turning left/right).
Historical Photos (12–18 photos): Genuine, older pictures of yourself from your personal gallery. These must span a timeline of at least 1 year up to 20 years to help train models on natural aging.
Crucial Quality Standards (Submissions failing these will be rejected):
No Filters or Edits: Do NOT use portrait mode, beauty filters, skin smoothing, or any photo editing software. All photos must be completely raw and natural.
Original Metadata (EXIF): Your selfie files must contain their original camera data (date, time, and device details). No screenshots or scans of printed photos are allowed.
Clear Face Visibility: No sunglasses, masks, hats, or hair blocking your eyes, nose, or mouth.
Technical Specs: Images must be in JPEG format, with a minimum resolution of 512x512 pixels for the face area, and file sizes between 128 KB and 300 MB.
Payout & Compensation:
High payout rates are offered for every fully validated and accepted image dataset (up to 24 billable images per participant). If you are interested, please apply with your location and smartphone model. Rate is negotiable.
Apply on Freelancer →
Project sourced from Freelancer.com. Applications happen directly on the original platform — we never collect your data.