MIL
NEURHA:
Signal Processing Platform

Platform includes Model Zoo for wearables ML tasks that allows to get a production-ready solution in several days

Flexible signal processing methods allow to get data- and human-specific solutions

Better and faster decision making in processes based on activity and sequences of activity analytics
Improving user experience while product usage by integrating wearables-processing features
BUSINESS CASES
1
HAR, gestures recognition by wearables signal
2
PDR in navigation and SLAM for position tracking
3
Vibration diagnostics and anomalies search
4
Forecasting of business indicators and metrics
SOLVING METHODS
Signal Segmentation, Classification, Regression, Clustering for IMU, PPG, other wearables
Signal augmentation and labeling
Transfer learning in signal processing problem statements: decoding, segmentation..
Physical & technical models for models accuracy improvement and decreasing of model size
EXPERIENCE & EXAMPLES
Made on
Tilda