In developing novel object detection, tracking and recognition solutions we focus on machine learning and stereo vision technologies. These technologies are combined and customized according to the specific requirements. In the field of machine learning we apply deep learning for various applications, such as object recognition for video surveillance, logo detection for critical content analysis, and text detection for media analysis. In the field of Augmented Reality (AR) we develop model-based tracking algorithms which run on mobile devices in real-time. Stereo vision technologies significantly increase the object detection and tracking reliability. Typical applications include people tracking over various camera sensors, people counting in crowded scenes, and queue length estimation.