Demographics
Anonymous age and gender estimation, staff segmentation, and how the AI is tuned.
Anonymous age and gender estimation, staff segmentation, and how the AI is tuned.
Glossary
- Age band An age band is one of the seven bracketed categories Aura Vision uses to classify a visitor's estimated age — coarser than an exact age, but reliable, useful and privacy-friendly.
- Gender estimation Gender estimation is the anonymous classification of each visit as male or female based on body shape, gait and clothing — never facial recognition — with results reported only as aggregate counts.
- Staff segmentation Staff segmentation is the process of separating employees from customers in the count by training the AI to recognise a distinctive visual cue — typically a uniform, polo, apron or lanyard.
Questions
- How does it recognise age and gender? We use an AI-powered algorithm that has been trained on massive datasets.
- How does it recognise staff? We train the algorithm to recognise a visually distinct cue such as staff uniform or a lanyard using images provided by you.
- How accurate is gender recognition? Gender recognition is 95% accurate on average across male and female categories, measured against ground-truth audits at deployment.
- How accurate is Age recognition? Age recognition is 85–90% accurate across seven age bands, measured against ground-truth audits at deployment.
- How accurate is Staff recognition? Staff recognition is 99% accurate when employees wear the distinctive visual cue the AI was trained on — typically a branded uniform, polo, apron or lanyard.
- Does it count infants? Infants are counted in the Under 16 demographic group. Aura Vision groups all under-16s together rather than estimating an exact age for children.
- Can we choose custom age groups? The age group segmentation has been trained on massive data sets and these cannot be amended.
- Can I remove an age-group category? Yes, it is possible to remove a single age group category. We could, for example, remove the under-16 age group category.
- Are there any measures in place to ensure demographic estimates are not biased or discriminatory? Aura Vision starts with a trained generic AI model that is fine-tuned per store to improve accuracy and reduce bias.
- Can demographics be switched on for individual stores or groups of stores? Yes. Demographics default to an organisation-wide setting, but stores can be split into groups with different demographic configurations if needed.