In the recently held "2016 Asian consumer electronics show (CES Asia 2016)", sogou map issued its first full voice "sogou intelligent navigation".
Sogou map Kong Xianglai, general manager of the securities journal reporter interview, "the emergence of intelligent navigation sogou, means that we have not just simply stay on the phone, but to cover all aspects of the intelligent hardware. In the future, sogou map in the Internet, smart wear the layout of the equipment will be more".
As the industry's tendency towards intelligence, automation, human-machine interaction potential applications began to highlight.
It is reported, this time to focus on car networking human-computer interaction and integration of third party content service platform sogou maps, POI to the evolution of the POP, by means of techniques make maps the car service with a link to the real world outside the entrance. The technology using sogou's own voice technology and semantic understanding, refactoring to the data, realize the whole speech interactive driving scenarios.
Booth design with the help of a polyhedron of urban structure, sogou map shows people its speech recognition and semantic understanding technology and data depth reconstruction "sogou intelligent navigation". According to introducing, as a Hand - free&Eye - free speech interactive products, it can provide users with from search to navigate to the intelligent, integrated voice services. Can run the mobile phone already, also can through the Internet protocol to connect to the machine, or run on the machine operating system directly.
At the same time, sogou map will own voice technology and applied to the intelligent navigation "sogou" semantic understanding, make navigation become more "sensitive". Speech technology, in view of the POI optimized recognition accuracy reaches 99%, after 4000 hours of car voice training, customize for car environment, with inhibition of wind noise, echo cancellation, its advantage of filtering and phonetic error correction. Semantic understanding, to realize the man-machine dialogue, support more than 100000 nodes of semantic web, heuristic dialogue model to ensure the convergence degree of session as high as 80%.
