Segmentive is a breakthrough video segmentation technology for mobile empowering users to intelligently remove themselves from live backgrounds to be transported into a scene of their choice, creating impossible moments in real-time. CUTTING-EDGE VIDEO SEGMENTATION TECHNOLOGY FOR MOBILE • Despite attempts by major technology platforms, the creation of high-quality green screen style video on smartphone has been unachievable. • Even the most advanced mobile processors lack the architecture and power required, resulting in low -quality visuals that are characterised by unclear edges and disappearing subjects. TECHNOLOGY • Segmentive’s technology combines deep learning techniques with advance neural network algorithms to replicate human understanding of scene identification, on mobile and in real-time. • Deep Neural Network >> Pre-classification of each frame into background and foreground pixels uses a 92-layer, pre-trained Deep Neural Network (DNN). • Generalised Learning Vector Quantization (GLVQ) >> Fine-classification uses GLVQ algorithms to identify and update the characteristics of each frame in real-time which also autonomously learn during run time. • Third Party Information >> Our technology also allows and utilises third party information such as depth sensor from the camera lens to create additional parameters, enhancing the quality of the Segmentation. • Composition >> Software automatically weighs and combines pre-classification and fine-classification by splitting the frame into areas based on classification confidence and visual characteristics. RESULTS The ultimate mixed reality technology for smartphones, Segmentive enables users to seamlessly fuse physical and digital worlds in real-time to create innovative and unique video content that can be instantly shared or streamed in any format. GET IN TOUCH TODAY We are only an email, phone call and Whatsapp away.
