The Synchronous Data Stream (SDS) Framework implements a data stream management, provides methods and helper tools for developing and optimizing embedded applications that integrate DSP and ML algorithms. This framework relates to the Compute Graph streaming that is part of the CMSIS-DSP compute library.
-
Implements a flexible data stream management for sensor and audio data interfaces
- Supports data streams from multiple interfaces including provisions for time drifts.
-
Provides methods to record real-world data for analysis and development
- Input to Digital Signal Processing (DSP) development tools such as filter designers
- Input to Machine Learning (ML) model classification, training, and performance optimization
- Verify execution of DSP algorithm on Cortex-M targets with off-line tools
-
Playback real-world data for algorithm validation using Arm Virtual Hardware
- Input to Digital Signal Processing (DSP) development tools such as filter designers
- Input to Machine Learning (ML) model classification, training, and performance optimization
- Verify execution of DSP algorithm on Cortex-M targets with off-line tools
-
Defines binary data format with YAML metadata file.
-
Python-based utilities for recording, playback, visualization, and data conversion
The SDS Framework interfaces implement the SDS streaming interfaces via different methods. The functions of the sds.h Circular Buffer Handling may be used at any level of a Compute Graph to provide inputs or capture outputs of an DSP or ML algorithm.
The Sensor interfaces provide the software layer for the connection to the physical sensors.
Directory | Description |
---|---|
demo | Example implementations for various evaluation boards |
documentation | User documentation of the SDS Framework |
schema | Schema for SDS File Format |
sensor | Sensor interface for the SDS Framework |
sds | Interfaces of the SDS Framework for Cortex-M devices |
utilities | Python scripts for processing of SDS binary data files |