PERCEPT / TECHNICAL EVALUATION
Custom AR tracking integration
When a standard SDK does not fit your camera pipeline or target requirements, the integration contract matters as much as the tracking algorithm. Work with Holoceive to scope Percept evaluation and project-specific engine integration.
Choose the tracking problem before choosing a wrapper
Image target tracking, object pose tracking and device motion tracking solve different problems. Describe what needs to remain aligned and relative to which coordinate frame. Percept discussions currently focus on image and object tracking. Web and VIO requirements can be reviewed, but should not be treated as committed production SDK features.
Write down the camera and timing contract
Specify frame format, resolution, orientation, timestamps, camera intrinsics if available, and ownership of capture and buffers. Explain whether another AR framework already controls the camera. Agree on how tracking results are timestamped and transformed into application coordinates, and how the application reacts to missing frames or tracking loss. Public API names and binary interfaces have not been finalized.
Plan the Unity and native lifecycle
For Unity, share the version, rendering pipeline, build targets and existing native plugins. For iOS and Android, include minimum OS and device requirements, build system constraints, camera permissions and suspend/resume behavior. Threading, resource cleanup and packaging should be part of the agreed integration deliverable. Support for your exact combination must be confirmed in evaluation.
Separate evaluation from production licensing
A useful first scope identifies the target, devices, sample data, test method and success criteria. Production discussion can then address app distribution, updates, support, target preparation and any custom work. Percept standard pricing and licenses are not finalized. Evaluation costs, delivery format and customization fees are agreed before paid implementation begins.
Considering a Vuforia or VisionLib alternative?
Bring the concrete limitation that led you to look elsewhere: target preparation, camera access, deployment constraints, tuning requirements or commercial terms. Compare candidates against the same acceptance tests and current vendor terms. Percept does not claim superior speed or accuracy without a comparable measurement. Use the SDK alternatives page to structure that discussion.
SDK alternatives ↗What to include in your evaluation request
- Tracking target and required output coordinate system
- Camera owner, frame pipeline and timing constraints
- Unity version or native build environment and target devices
- Offline, network, data handling and distribution requirements
- Evaluation criteria, expected maintenance and delivery timeline
Project-based integration with the Percept engine team.