Face processing
How fast can we recognise a famous face?
Open access · cc by · source: Europe PMC
Even under extreme time pressure, people need at least about 360-390 milliseconds to recognise a famous face among unfamiliar ones, much longer than simply detecting a face.
Study at a glance
- Design
- Human experiment — Speeded go/no-go famous-versus-unknown face task; Experiment 1 used strong speed encouragement, Experiment 2 added a 600 ms response deadline with audio feedback (the SAB procedure); minimal RT estimated from when hits first reliably outnumber false alarms
- N
- Experiment 1: 31 volunteers. Experiment 2: 101 recruited, 14 failed the task leaving 87, of whom a minimal RT could be computed for 57
- Population
- Healthy young adult volunteers with normal or corrected-to-normal vision
- Outcome
- Minimal reaction time (minRT) for correct famous-face responses, median RT, d-prime sensitivity and response bias C
Structured fields used in claim comparison tables when every cited study has a complete layer.
Key findings
In Experiment 1 the median minimal RT across people was 555 ms, but the two fastest participants reached 390 ms with normal accuracy. In Experiment 2 the median minimal RT dropped to 420 ms and the fastest individual was 360 ms (350 ms pooled across all trials). Faster people were also more accurate in Experiment 2, so there was no speed-accuracy trade-off, and participants responded less conservatively than in Experiment 1. The authors conclude recognition cannot happen faster than about 360-390 ms.
Methodology
In Experiment 1, 31 people saw photos flashed for 100 ms and lifted their fingers as fast as possible only when the face was famous, with feedback and encouragement to beat their times. Experiment 2 tested 101 people with a stricter 'speed and accuracy boosting' procedure: responses had to come within 600 ms and each answer received immediate positive or negative audio feedback. The key measure was each person's minimal reaction time, the earliest point at which correct 'famous' responses significantly outnumbered false alarms.
Limitations
The two experiments used different stimuli (whole photos in context versus cropped grey faces chosen as the most recognisable), so their results cannot be compared directly as a pure test of the procedure. It measures familiarity for celebrities only, not personally known or newly learned faces, and the task can be done on a feeling of familiarity without full identification. The neural interpretation about the perirhinal cortex and feedback processing is inferred from reaction times and prior literature, not measured in this study. In Experiment 2, 14 people failed the task and a minimal RT could not be computed for many others, so the estimate rests on a subset.
How this study connects
Role on claims
Each row is a claim on a concept or method page where this paper supports, challenges, or qualifies the statement. Roles are hand-checked — not a model guess.
Not yet placed on a claim. This paper has study layers, but no concept page yet cites it as support, challenge, or qualifier.
Related papers in this topic
Same topic cluster — not a recommendation engine.