Gopher Protocol Inc. (OTCQB: GOPH) (“Gopher”), a company specializing in the creation of Internet of Things (IoT) and Artificial Intelligence enabled mobile technologies, including a global platform with both mobile and fixed solutions, is pleased to announce it started the development of Avant! AI Deep Neural Network (DNN) for advanced object detection. Object recognition is a computer vision technique for identifying objects in images or videos. Object recognition has become a key feature of deep learning and machine learning algorithms in the past decade. As we move forward with Avant! AI Engine development, a comprehensive image analysis is needed in order to detect and identify objects as having more precise and detailed object recognition is a crucial task especially for autonomous driving, anatomical/facial recognition and robotics. Avant! will need to precisely define an object’s geometry in order to classify and identify them. In addition, it needs to locate objects contained within images, which is called object detection process. Gopher started the development of DNN, a machine learning based system, in order to achieve a prominent state-of-the-art object detection system. The system will capture object’s parts, ratios and geometrical relations.
“The main challenges in objects detection is to define the representations and kinematics of the objects, expressing them as graphical models. This functionality will be achieved by our machine learning DNN (Deep Neural Network) model for Avant! AI,” stated Danny Rittman, Gopher’s Chief Technology Officer. “Our Deformable Part-based system will consist of broad class of proprietary detection algorithms, used on images in order to achieve efficient classification and objects recognition,” Dr. Rittman continued. “Using discriminative learning technique of graphical models allows for building high-precision part-based models for wide variety of object classes. We expect that our deep learning architectures will have the capacity to learn more complex models and produce robust training algorithms for Avant!. When fully developed, these architectures are expected to allow Avant! to recognize objects without the need for human intervention. Upon development, we expect the technology to provide a state-of-the-art technology for Avant! AI, enabling it to “see”, locate, classify and identify objects, which we believe it will use in a wide variety of Gopher’s future applications.”