Skip to main content

Project Aria Machine Perception Services

To accelerate research with Project Aria, we provide several Spatial AI machine perception capabilities that form the foundation for the future Contextualized AI applications and analysis of the egocentric data. These capabilities are powered by a set of proprietary machine perception algorithms designed for Project Aria glasses and provide superior accuracy and robustness on recorded data compared to off-the-shelf open source algorithms.

All MPS functionalities are offered as post-processing of VRS files via a cloud service. Use the Desktop App, to request derived data from any VRS file that contains necessary sensor's data.

Current MPS offerings

The following MPS can be requested via the Aria Desktop app for data you've recorded with a Project Aria device, as long as the data has been recorded with a compatible Recording Profile. The Recording Profile Guide provides a quick list of compatible sensor profiles or go to Recording Profiles in Technical Specifications for more granular information about each profile.

6DoF trajectory

MPS provides two types of high frequency (1kHz) trajectories

  • Open loop trajectory that is a local odometry estimation from visual-inertial odometry (VIO)
  • Closed loop trajectory that is created via batch optimization, using multi-sensors' input (SLAM, IMU, barometer, Wi-Fi and GPS), fully optimized and providing poses in a consistent frame of reference.

Request trajectory (location data) in the Desktop app to get these outputs, the recording profile must have SLAM cameras + IMU enabled.

Online sensor calibration

The time-varying intrinsic and extrinsic calibrations of cameras and IMUs are estimated at the frequency of SLAM cameras by our multi-sensor state estimation pipeline.

Request trajectory (location data) in the Desktop app to get these outputs, the recording profile must have SLAM cameras + IMU enabled.

Semi-dense point cloud

Semi-dense point cloud data supports researchers who need static scene 3D reconstructions, reliable 2D images tracks or a representative visualization of the environment.

In the Desktop app, this can be requested as an addition to trajectory (location) derived data and has the same recording profile requirements.

Eye gaze data

Eye gaze is the most important indicator of human’s attention, eye gaze direction estimation with uncertainty is provided by MPS. Eye gaze estimation uses the data from the Eye Tracking (ET) cameras.

Request Eye Gaze data in the Desktop app to get these outputs, for any recording that had ET cameras enabled.

If you have made a recording with In-Session Eye Gaze Calibration, you will receive a second .csv file with calibrated eye gaze outputs.

About MPS Data Loader APIs

Please refer to our MPS data loader APIs (C++ and Python support) to load the MPS outputs into your application. Additionally, the visualization guide shows how to run our rich visualization tools to visualize all the MPS outputs.

Questions & Feedback

If you have feedback you'd like to provide about overall trends and experiences or improvement ideas we'd love to hear from you. Please email AriaOps@meta.com or post to the Project Aria Academic Partner Feedback and Support group.