Every interaction has a rhythm.
The demo looks at timing and movement during this test.
HUMAN CONTROL / DEMO
Human or agent?
A few interactions. An estimate of what’s behind them.
1 Start a test2 Type, click & scroll3 See the result
Ready to capture0.0 s
01 / INTERACT
Try it yourself
Move your mouse and click the three targets.
No target selected
Scroll this panel
Type naturally.
Use an invented sentence. Typed text and key identities are excluded from measurements.
Move between the targets.
Try a few clicks, then scroll back up. Choose Finish test to see your result.
A small sample, a useful signal.
Human-like agents can resemble people. This demo provides an experimental estimate.
02 / ASSESSMENT
ReadyYour result appears here
Start a test, complete the task, then choose Finish test for a full-session result.
MODEL ESTIMATE
Experimental. Human-like automation can be mistaken for a person. Scores are model estimates, not calibrated confidence.
Experimental browser-based detection. This demo estimates input origin, not a person’s identity.
Only this test area is measured. Typed text and key identities are excluded. Camera stays off unless you enable it.
Camera preview
Gaze preview
Centre your whole face in the preview. Adjust the camera until your forehead, eyes and chin are visible, then keep that position. The camera will stop after 30 seconds if you do not continue.
Quick mode uses nine calibration locations and five checks. Detailed mode uses 16 locations and nine checks, takes about 60–90 seconds and remains experimental. Processing stays in your browser. Gaze does not affect the human-or-agent result.
You can select this now. The point appears after calibration passes, without starting a capture. Gentle display smoothing reduces jitter with a small delay; accuracy checks and exported measurements use the unsmoothed estimates. Selecting this does not turn on the camera.
Calibration diagnostics
Off by default. If enabled, calibration video and diagnostics save on this computer. No audio or screen recording. Keep this page open until saving finishes.
Debug recording is off. The camera is off.
The latest numeric calibration report is saved in this browser until cleared. Downloads include pixel errors and camera settings, with no video, face landmarks or fitted weights.
Camera off. Gaze is optional.
Calibration quality
Not calibrated. Calibration is approximate and depends on lighting, camera position and how steadily you look at each dot.
No current gaze estimate.
If the first check fails, you can retry all check dots once. The camera waits for up to 30 seconds, with no gaze inference during that wait. Detailed mode warns when your head moves outside its calibration range. Finish, Clear, leaving this tab, resizing the window or Stop camera releases the camera; the last numeric report stays available. A passing coarse mapping check does not guarantee small-target precision.
Advanced options
Checking classifier availability…
Finish test assesses the full session.
No window has been assessed.
Full-session assessment is the default. Enable Sliding window for updates during the test. Free interaction stops after two minutes; guided tasks allow up to five minutes.
Extra browser measurements are collected separately and do not affect the current classification.
The optional probe observes selected Element method calls inside the task. It excludes arguments and DOM contents and cannot see every inspection path.
For research only. Labels are saved with downloads and never influence predictions. Use a code, not a name.
Test details & downloads
LATEST WINDOW
Test details
Downloads contain timing and pointer measurements, with no typed text or key identities.
Start a test to see its measurements.
Evidence captured
Detection sample requirements are not yet validated. Calculable measurements alone do not establish enough evidence to identify input origin.
LOCAL CLASSIFIER RESPONSE
Raw response
No response yet.
No response yet.
Recent windows
Each row describes its recorded interval. Overlapping windows share evidence and are not independent tests. The latest 60 outcomes are retained.
| Window | Result | Human / automation | Input | Latency |
|---|
Additional Behavior Measurements
No additional measurements yet.
Unavailable values reflect missing, unsupported or boundary-cut evidence. These measurements do not identify the operator and are not trained into a classifier yet.
Group and measurement names explain each value. Hover over a measurement name to see its original data key; the full keys also remain in the downloaded summary.
Inspect the 74 additional measurements
Gaze measurements for this window
No gaze summary for this result yet.
Measurement downloads contain numeric gaze summaries. Debug recordings are separate local video and diagnostic files when enabled above. Facial landmarks and calibration weights are never saved. Missing gaze is not evidence of automation.
Gaze–action coordination
No gaze–action measurements for this result yet.
These measurements connect gaze available before each action to its selected target. The tolerance around a target comes from calibration error; nearby targets may be ambiguous. A person can watch an agent act, so target alignment does not identify the executor. These features are collected for future classifier training and do not change the current classification.
Inspect recent actions (up to 20 shown)
Inspect the feature summary
No snapshot yet.