We recently conducted research in Heterogeneous Data Discovery Using Deep Neural Networks as part of the Small Business Technology Transfer Research (STTR) program with the Air Force Research Laboratory (AFRL). The effort leveraged our expertise in combining machine learning with advanced signal processing techniques for developing processing solutions, and focussed on the development of novel techniques to jointly process collected data from sensors of varying types. KickView continues to utilize this technology enhance our multi-sensor analytics products and technology offerings.
CNNs can be utilized for low-SNR signal detection problems. We used a custom CNN trained on noisy spectrograms to detect events in RF signal data. The network has an impressive capability of detecting signals below the noise floor without knowing the start or stop time/frequency, which is difficult using traditional signal detection techniques.
We created an introductory lab for the NVIDIA QwikLabs site. Go check it out!
At KickView we feel fortunate to be living and working in beautiful Colorado. There are many outdoor activities to enjoy and the mountains provide close access to a variety of recreational opportunities. There is also a thriving technology community, with frequent meetups and conferences to participate in. Whether you enjoy working at a startup or a large corporation, Colorado has it all. If you find yourself in the Denver area and would like to learn more about KickView, send us a message! See our career page for our current career opportunities – Kickview.com/careers