Winkam Lab

WINKAM Lab was growing and receiving good customer feedback, but due to COVID-19, the business environment became too challenging to sustain the company, so we decided to gradually wind down and let our employees find other jobs. Our company was dedicated to crafting C/C++ libraries tailored to distinct hardware and application scenarios within the Machine Learning domain, encompassing real-time Pattern Recognition and Computer Vision. Our proficiency spanned across diverse sensor categories, including MEMS accelerometers, gyroscopes, UWB XYZ, and video cameras. A substantial portion of our endeavors were centered on real-time sensor data, particularly concerning the human body's interactions. For instance, we harnessed sensor data to facilitate gesture control on mobile devices, detect running technique anomalies via accelerometers and gyroscopes to prevent injuries, ascertain jumpshot release timings with millisecond accuracy using XYZ UWB systems on basketball courts, and enhance road safety by identifying cyclists through mono cameras. Our overarching objective encompassed achieving unparalleled precision in real-time signal data recognition. Our team developed specialized perception (pattern recognition) libraries in C/C++, seamlessly integrated into embedded software for wearable and IoT enterprises. We conducted expansive research and development initiatives on a global scale, collaborating on scientific undertakings with prominent entities such as Intel, Specialized Bicycles, Stanford, and KIT.