Download technical_boundaries.md from origin-data-lab/video-imu-capture-evaluation: direct link, hf CLI and curl.
- Browser
- Download file 1.54 kB
-
https://huggingface.co/datasets/origin-data-lab/video-imu-capture-evaluation/resolve/main/technical_boundaries.md
- Command line
-
hf download hf://datasets/origin-data-lab/video-imu-capture-evaluation/technical_boundaries.md
-
curl -L -o technical_boundaries.md https://huggingface.co/datasets/origin-data-lab/video-imu-capture-evaluation/resolve/main/technical_boundaries.md
Technical Boundaries
The measurements in this public evidence package describe one tested device and software configuration during a continuous real-world field recording.
This Evaluation Does Not Claim
- Hardware frame-level video-IMU synchronization
- Sub-frame synchronization
- Camera-exposure-level timestamp synchronization
- A guaranteed 410 Hz hardware specification
- Guaranteed continuous 100 Hz application-level delivery
- 12 independent experimental sessions
- A determined cause for the observed timing gaps
- Universal performance across every device, environment, capture architecture, or future project
Observed Limitations
One of 12 evaluated segments showed a maximum video inter-frame gap above 100 ms.
The maximum observed video gap was 499.80 ms.
Application-level accelerometer and gyroscope streams showed gaps above 100 ms in all 12 evaluated segments.
Maximum observed gaps were:
- Accelerometer: 589.34 ms
- Gyroscope: 599.08 ms
Native IMU streams did not show gaps of comparable magnitude in this evaluation.
The cause of the application-level timing gaps has not been established.
No attribution is made to buffering, batching, Android scheduling, downsampling, resampling, or another specific mechanism.
Interpretation
The evidence supports measured native sensor timing characteristics, timestamp integrity, and capture-session temporal bracket coverage for the evaluated configuration.
It does not support stronger synchronization or universal hardware-performance claims.