How EchoPitch Uses FACS to Measure Your Confidence
Not AI hype. Fifty years of peer-reviewed science, applied to your presentations.
Quick Answer
EchoPitch measures your confidence using the Facial Action Coding System (FACS), the scientifically validated framework developed by psychologists Paul Ekman and Wallace Friesen at UC San Francisco in 1978 that defines 44 Action Units, each tied to a specific facial muscle — for example AU6 (orbicularis oculi) for a genuine smile or AU4 (corrugator supercilii) for concentration or concern. When you record a practice session, EchoPitch’s EchoDepth AI tracks all 44 Action Units frame by frame in real time, detecting which are active, at what intensity, and for how long. These patterns are mapped onto the Valence-Arousal-Dominance (VAD) model, with Dominance being most critical for presentation coaching: low Dominance signals feeling controlled or submissive, which is the primary marker of presentation anxiety and what audiences read as a lack of confidence. Because most people misjudge their own expressions — feeling far more nervous than they appear — EchoPitch flags the exact moments your confidence signals drop, rather than giving vague feedback like 'you seemed a bit nervous.'
MBA, Bayes Business School · Founder, Cavefish · May 2026
What Is FACS and Why Does It Matter?
The Facial Action Coding System (FACS) is the most comprehensive and scientifically validated framework for describing human facial movement. Developed by psychologists Paul Ekman and Wallace Friesen at the University of California San Francisco in 1978, it has been used in clinical psychology, law enforcement, animation, and — since the 2010s — AI emotional analysis.
FACS defines 44 Action Units (AUs), each corresponding to a specific facial muscle or muscle group. AU6 is the orbicularis oculi — the muscle around the eye that activates in a genuine smile. AU4 is the corrugator supercilii — the brow muscle that signals concentration or concern. Every human emotional expression is a combination of these units.
How Does EchoPitch Track Action Units During Your Sessions?
When you record a practice session with EchoPitch, Cavefish’s EchoDepth AI analyses your facial video in real time — tracking all 44 Action Units frame by frame. The system detects which AUs are active, at what intensity, and for how long during each section of your presentation.
These AU patterns are then mapped onto the Valence-Arousal-Dominance (VAD) model — a three-dimensional framework for emotional state. For presentation coaching, the Dominance dimension is most critical: low Dominance indicates feeling controlled or submissive — the primary signal of presentation anxiety — and is what audiences read as a lack of confidence.
What Does This Tell You That You Cannot See Yourself?
Most people dramatically misjudge their own emotional expression. People who feel nervous often look less nervous than they think. But specific AU patterns — micro-expressions, brow tension, tight lip corners — reliably signal anxiety to trained observers and, it turns out, to audiences.
EchoPitch identifies these patterns and flags the specific moments in your presentation where your confidence signals drop. This is far more actionable than "you seemed a bit nervous" — it tells you exactly when, and which aspects of your delivery changed.
EchoPitch is powered by Cavefish’s EchoDepth platform — the same emotional AI used in enterprise deployments including financial services vulnerability detection, defence sector reliability assessment, and elite sports performance monitoring. Learn more about the science →
See Your FACS Confidence Score
Record a practice session and see how your confidence signals change moment to moment. 5 sessions, paid once. No subscription.
Try EchoPitch →Get daily tips on Reddit
Join our community for presentation tips, Q&A, and support.