Latin American Egocentric Video Dataset

I need a dependable data-collection partner who can capture real-world, first-person video inside Latin American retail stores. Every clip must show genuine shopping behaviour from the wearer’s point of view so the resulting dataset can be used to train computer-vision and robotics models in activity understanding.

Scope
• Film exclusively in retail environments—no restaurants, offices, commuting, or cooking footage needed.
• Capture shopping activities: browsing aisles, picking items, using self-checkout or staffed counters, queuing, paying, bagging, exiting, and similar moments that naturally occur during a store visit.
• Shoot in multiple countries or cities when feasible to cover different cultural layouts, product ranges, lighting conditions, and customer demographics.

Technical expectations
• First-person, head-mounted or chest-mounted cameras (e.g., GoPro, Insta360) at 1080p/30 fps or better, natural audio on.
• Continuous takes preferred over short bursts to preserve action context.
• Avoid staged or scripted scenes; authenticity is critical.

Deliverables
1. Raw video files organised by location and date.
2. A lightweight metadata sheet (CSV/JSON) noting store type, city, timestamp ranges, and basic participant demographics (age band, gender, handedness).
3. Signed consent forms from all identifiable participants, compliant with local privacy regulations.

Acceptance criteria
• Minimum 10 hours of usable footage delivered per milestone.
• ≤5 % footage rejected for blurriness, severe occlusion, or privacy violations.
• Metadata and consent documentation must align with the video batch and pass a quick manual audit.

If you already have Latin American retail access or previous egocentric-capture experience, let me know. Please outline your proposed collection plan, countries you can cover, sample equipment, and timeline so we can move forward quickly.