Monado

This page shows a camera–IMU calibration of the Monado sequences (TUM MIC). It is the dataset where two checks earn their keep:

t_ga: the time offset between the gyroscope and accelerometer channels
Round-trip validation: we apply the result to its own data and calibrate again

The data comes from a Valve Index headset, not a research rig. It ships with Monado’s own published calibration. So it gives a second opinion on TUM VI from a different toolchain.

The headset has a sensor page with its specs, solved rig and results.

The rig

Sensor	Detail
2 × monochrome cameras	Stereo fisheye
IMU	`/imu0`
Reference	Monado / Valve Index, baseline 134.0 mm

Board: AprilGrid 6×6, tag_size 0.03 m, tag_spacing 0.3, family t36h11.

Getting the data

The data uses the EuRoC mav0 layout, so no conversion is needed:

tum_mic/mic_calib/MIC05_imucalib2/mav0/

Main sequence: MIC05
Also configured: MIC01, 02, 03, 07, 08, and 11

The extra sequences let us measure the spread across sequences.

Configuration

input:
  format: euroc
  path: /data/tum_mic/mic_calib/MIC05_imucalib2/mav0
  imu_topics: ["/imu0"]
  window_mode: motion        # auto-pick the most-exciting 80 s
  segment_length: 80.0
  cache_dir: /tmp/mvis_detcache

board: {board_type: apriltag, rows: 6, cols: 6, tag_size: 0.03, tag_spacing: 0.3, tag_family: t36h11}

camera_input:
  path: examples/regression/tum_vi_mic/MIC_camchain_v2.yaml

solve:
  max_iterations: 40
  init_kwargs: {keyframe_hz: 2.0, scale_rel_tol: 0.3}
  estimate_t_ga: true
  sigma_t_ga: 0.010

estimate:
  camera: {intrinsics: false, tr: false, td: true, extrinsics: true}
  imu:    {intrinsics: dw_da, model: calyx}

output:
  path: results/regression/tum_vi_mic/mic05/cam_imu.json

reference:
  label: Monado Valve Index
  baseline_mm: 134.0

Two choices to note:

window_mode: motion picks the most-excited 80 s for you. You do not have to hunt for it.
Intrinsics are pinned to the committed camchain. The solve then focuses on extrinsics, td (the camera–IMU time offset) and the IMU.

Running it

mvis cam-imu -c examples/regression/tum_vi_mic/cam_imu_mic05_tga.yaml

Validation — round-trip

We do not accept the result on its residual alone. We apply it to its own data and calibrate again:

mvis roundtrip -c examples/regression/tum_vi_mic/cam_imu_mic05_tga.yaml \
  --bag MIC05.bag

Quantity	Before	After correction	Verdict
`Da`	1.06	1.00	collapsed ✅
`b_a`	−0.25	−0.01	collapsed ✅
Stereo baseline	133.8 mm	133.8 mm	preserved ✅

Both checks pass at once, and they pull in opposite directions:

IMU terms collapse: the correction absorbs the accelerometer scale and bias that the first pass claimed.
Geometry holds: the stereo baseline does not move.

Monado publishes a 134.0 mm baseline. Our 133.8 mm is 0.2 mm off.

See Round-trip validation.

`t_ga` on and off

We configure both variants:

cam_imu_mic05_tga.yaml estimates t_ga
cam_imu_mic05_notga.yaml does not

This isolates the effect of estimating the gyro↔accel offset on real data, rather than arguing about it. See IMU model. See also Why calibrate t_ga.

Reproducibility

Item	Value
Configs	`examples/regression/tum_vi_mic/`
Sequence	`MIC05_imucalib2`, motion window 80 s
Board	AprilGrid 6×6, 0.03 m, spacing 0.3, `t36h11`
Camera intrinsics	pinned to `MIC_camchain_v2.yaml`
IMU model	`calyx`, `dw_da`, `t_ga` estimated
Reference	`monado_reference_calibration.yaml`, baseline 134.0 mm