AI Facial Data Collection Contributor

$139–140/hrrate as listed

Contribute recorded facial imagery and short video clips of yourself, under a signed consent and likeness agreement, to build training and evaluation sets for face-related AI systems. Work is session-based and protocol-driven: you follow precise instructions on framing, lighting, expression, angle, and occlusion, then submit takes for QA review.

Summary, eligibility guidance and screening preparation are prepared by Sidequest from the platform's posting. Check the source listing before applying.

What the work involves

This is a data contribution role rather than an annotation or red-teaming role. micro1 recruits people to sit for structured capture sessions — typically a smartphone or webcam recording in a room you control — and produce a defined set of takes: neutral face at fixed angles, scripted expression sequences, movement passes, varied distances, and often occlusion conditions such as glasses, masks, hats, or partial hand coverage. Some briefs add lighting variation (window light, overhead, low light) or require multiple sessions across different days and outfits so the dataset captures natural variation.

Each brief comes with a spec sheet and reference examples. The practical skill is compliance and consistency: correct resolution and frame rate, stable framing, no background people, no compression artifacts, files named and uploaded exactly as instructed. Rejected takes are usually rejected for mundane reasons — motion blur, clipped chin, mixed color temperature, background clutter, audio when audio wasn't requested. Contributors who read the spec carefully and re-shoot before submitting clear QA far more often.

What the platform screens for

Verified identity and legal capacity to consent. Expect government ID verification and a signed likeness/data-use agreement covering scope, retention, and permitted downstream use. Read it. Facial data is biometric data, and the terms are the substance of the deal.
Demographic and appearance fit to an open brief. Most collections are quota-based: a specific age range, skin tone range, eyewear status, facial hair, or region. Fit is the gate, and it is nobody's fault when a brief closes.
Equipment and environment. A recent phone or a 1080p+ webcam, a room you can control, and reliable upload bandwidth.
Reliability across sessions. Multi-day collections fail when contributors disappear after session one.

Logistics and pay

Fully remote and asynchronous — you shoot on your own schedule within a submission window. Individual sessions are short, often 20–60 minutes, and briefs may recur or may be one-off. The $139–140/hr band is as observed on the listing; in practice this kind of work is frequently paid per completed and QA-approved session or per accepted asset, so the effective hourly rate depends on how much unpaid setup and re-shooting you absorb. Treat the band as an upper reference point, not a guaranteed wage, and confirm the payment unit before you commit to a multi-session brief.

What they look for

Selfies
Video Recordings
Attention to Detail
Follow Instructions

Before you apply

Would you pass this screen?

Drawn from what this listing actually requires. Answered in your browser — nothing is sent to us, and we do not store your answers.

Do you have a current government-issued photo ID (passport, driver's license, state ID, or equivalent) that you are willing to provide for identity verification as part of the platform's onboarding?
Government-issued ID · required
Do you have access to a smartphone with a camera or a webcam capable of recording at 1080p or higher resolution?
Video recording equipment · required
Do you have a room or space that you control (not a public place or shared space where other people appear on camera), where you can record facial video with a neutral or plain background?
Controlled recording environment · required
Do you have reliable internet bandwidth sufficient to upload video files (typically 100–500 MB per session) within a submission window of 3–7 days?
Reliable internet upload · required
Are you able to commit to completing all sessions in a multi-day brief if you accept one, without the likelihood of disappearing or canceling partway through?
Availability for multi-session briefs · preferred

0 of 5 answered

Included with this listing

Interview prep — 18 questions

Matched to what this platform's screening actually measures. Answer each out loud in 60–90 seconds: claim → concrete example → implication.

What it tests: Whether the candidate grasps that this role is about mechanical adherence to a spec sheet, not creative judgment, and can list concrete QA criteria (resolution, frame rate, framing, background, compression) rather than vague notions of 'good quality'.
A strong answer: A strong answer names at least three concrete spec items to verify — resolution, frame rate, file naming, background clutter, motion blur, framing of facial landmarks — and shows they understand rejected takes are usually mundane technical failures, not artistic ones. They should indicate they will re-check the spec before submission.
Where people lose it: Treating the role as subjective or beauty-focused ('I'll make sure I look good') rather than technical ('I'll verify I hit the exact resolution and frame rate in the spec'). Also, not mentioning that they will re-shoot rather than submit borderline takes.