AI Tool Suite Custom Designed For RF
The OmniSIG SDK is a tool suite that enables customers to rapidly curate their RF datasets, train state-of-the-art deep learning inference models for custom wireless sensing applications, and deploy these to edge sensing devices. OmniSIG SDK is furnished with DeepSig’s industry-leading baseline RF dataset for machine learning, including many widely deployed consumer wireless signals. Customers can also incorporate their own custom data, signals and signatures for training the AI sensor.
THE SDK FOR MACHINE LEARNING AND WIRELESS DATA
The OmniSIG SDK contains tools for rapidly sorting, labeling and curating RF data, training the OmniSIG Sensor model with labeled data, evaluating its performance, and then deploying the trained deep learning model into an OmniSIG runtime. This toolsuite is a market-first, enabling customers to custom-tune DeepSig’s highly tuned and market-leading deep learning models for signal detection and classification for their specific needs, RF signatures and applications.
Unlike other labeling suites, which are today primarily built for computer vision and imagery applications, OmniSIG SDK is designed specifically for signal processing on complex-valued RF sample data by engineers with decades of industry experience. It contains a number of specialized features to assist in working with large RF datasets that simply don’t exist anywhere else.
DATA MANAGEMENT, LABELING & CURATION
AI workflows are driven by data. DeepSig has built a comprehensive labeling tool, designed specifically for working with signal recordings. The OmniSIG SDK provides an easy to use method for visualizing your signal captures and labeling them for use in AI systems using automated, semi-automated and hand-tuning methods.
TRAINING AND VALIDATING A MODEL
The world of artificial intelligence is moving fast, and DeepSig’s machine learning research scientists are at the top of the field. One of the primary goals of the OmniSIG SDK is to make the latest advances in AI, incorporated into the OmniSIG Sensor, available to customers without requiring them to be machine learning experts.
Using either data sets provided by DeepSig, included with an OmniSIG SDK license, and/or custom signal captures, customers then train the OmniSIG deep learning model until it reaches the desired level of performance.
Using a single desktop-class GPU, training an OmniSIG model takes on the order of a few hours, depending on the required level of performance. Models can be exported from a live SDK training process at any point for testing and validation.
DEPLOYING WITH OMNISIG
Once trained, the custom model can be easily deployed with the OmniSIG Sensor for immediate use in wireless sensing systems. The output of the model is a metadata stream that can then be used to facilitate downstream systems and operators to build advanced AI-based systems.