Subject Calibration

Subject calibration captures a neutral standing pose for each subject. This allows the SDK to scale a 3D biomechanical model to the size of the subject, which is required for biomechanical analysis.

Creating or Selecting a Subject

Each subject needs a profile before calibration. Height and weight are required — they're used directly by the analysis algorithm. Name and birth year are optional and can be left anonymous if needed.

Swift

// Create a new subject
let params = SubjectParameters(
    name: "Anonymous",       // optional
    weight: 70,              // kilograms, required
    height: 175              // centimeters, required
)
let subject = try await client.createSubject(parameters: params)

// Or select an existing subject (eg, first one of the list)
let subjects = try await client.subjectList()
let subject = subjects.first

TypeScript

// Create a new subject
const subject = await client.createSubject({
  name: "Anonymous",   // optional
  birthYear: null,     // optional
  height: 175,         // centimeters, required
  weight: 70           // kilograms, required
});

// Or select an existing subject (eg, first one of the list)
const subjects = await client.subjectList();
const subject = subjects[0];

Python

from modelhealth import SubjectParameters

# Create a new subject
params = SubjectParameters(
    name="Anonymous",   # optional
    weight=70,          # kilograms, required
    height=175          # centimeters, required
)
subject = client.create_subject(params)

# Or select an existing subject (eg, first one of the list)
subjects = client.subject_list()
subject = subjects[0]

Neutral Pose Instructions

Before recording, instruct the subject to:

• Stand upright with feet pointing forward
• Keep a straight back with no bending or rotation at the hips, knees, or ankles
• Stand completely still until told they can relax
• Be fully visible by all cameras

Make sure there are no obstructions in any camera's field of view before hitting record.

Recording the Neutral Pose

Swift

try await client.calibrateSubject(subject, in: session) { status in
    switch status {
    case .recording:
        print("Recording neutral pose...")
    case .uploading(let uploaded, let total):
        print("Uploading \(uploaded)/\(total) videos")
    case .processing(let percent):
        if let percent = percent {
            print("Processing: \(percent)%")
        } else {
            print("Processing...")
        }
    case .done:
        print("Neutral pose captured!")
    }
}

TypeScript

await client.calibrateSubject(subject, session, (status) => {
  switch (status.type) {
    case "recording":
      console.log("Recording neutral pose...");
      break;
    case "uploading":
      console.log(`Uploading ${status.uploaded}/${status.total} videos`);
      break;
    case "processing":
      if (status.percent !== undefined) {
        console.log(`Processing: ${status.percent}%`);
      } else {
        console.log("Processing...");
      }
      break;
    case "done":
      console.log("Neutral pose captured!");
      break;
  }
});

Python

from modelhealth import CalibrationStatus, CalibrationStatusUploading, CalibrationStatusProcessing

def on_status(status):
    if isinstance(status, CalibrationStatusUploading):
        print(f"Uploading {status.uploaded}/{status.total} videos")
    elif isinstance(status, CalibrationStatusProcessing):
        if status.percent is not None:
            print(f"Processing: {status.percent}%")
        else:
            print("Processing...")
    elif status == CalibrationStatus.recording:
        print("Recording neutral pose...")
    elif status == CalibrationStatus.done:
        print("Neutral pose captured!")

client.calibrate_subject(subject, session, on_status)

Troubleshooting

Neutral pose capture failed:

• Subject may have moved — instruct them to stand completely still until the recording is complete
• Check that the subject is fully visible by at least 2 cameras
• Reposition cameras to capture both sides of the subject and minimize occlusion (requires recalibration of the cameras)
• If the issue persists, try recalibrating the cameras

Inaccurate limb lengths or bad kinematics in analysis:

• Subject may not have been in the correct pose — refer to the image above and repeat the neutral pose capture
• Camera calibration may be off — try recalibrating

Next Steps

Once subject calibration is complete, proceed to Activity Recording.