Detection capability
Posture & Behaviour
Body position and movement patterns are analysed over time to identify physical cues associated with the emotional states expressed through posture and movement.
Posture & Behaviour analyses visible body position and movement patterns to identify physical cues associated with different emotional states. It considers how the body is positioned and how that position changes over time rather than relying only on a single moment. Happiness can be reflected through relaxed shoulders and an upright stance. Anger can appear through clenched fists and a forward lean, while sadness can be associated with slouching and a lowered head. Surprise can be reflected through lifted shoulders. Fear can appear through freezing or leaning back, while disgust can involve torso twisting or physical withdrawal. These visible posture and movement patterns provide another signal for understanding behaviour and emotional expression within the wider platform.
Deepfake Detection is one of eight signals used by the Trust Factor Engine. It is considered alongside the other seven signals rather than being treated as the only basis for the overall result. The engine considers multiple signals together when assessing whether the video, voice, or face being evaluated can be trusted. A single signal does not independently determine the final verdict. How the weighting works
Next
See the whole engine
Capabilities are components. The platform page shows how the eight combine into one score.
See it score a real video
Bring your own footage. We will run it through the Adaptive Cognito Engine and walk you through the findings.