Teraki embedded device SDKs cover a wide range of sensor data: time series, video and 3D data. The SDKs are the building blocks to develop embedded applications on various types of low-powered, edge hardware.
Time Series SDK
Supports encoding in a lossy or lossless manner to enhance the accuracy rates beyond those observed on raw data.
Average SDK-size is less than 10kB and increases CPU-efficiency with factor 10x and more.
With Teraki’s codec and ROI-detection, customers can classify objects as sky, cars, person and faces.
With TOI-detection, events (e.g. Lane Change) can be identified and the acquisition rate of videos can be dynamically adapted.
Using ROI and TOI customers’ AI-models will deliver a 10 - 30% improved object detection and classification rate per frame.
High performance at a low processing power (typically embedded ASIL CPU).
3D Point Cloud SDK
Reduction of 3D point cloud data in combination with lightweight segmentation.
Comparing with state-of-the-art products, this provides 10x to 150x better inference speeds and a 10 minute instead of a 3 days training.
The 3D point cloud SDK supports sensor fusion and SLAM.
Contact us for questions or trials: firstname.lastname@example.org.