Mobvoi independently researches and develops a full set of speech recognition technology, including online and offline speech recognition, and offline hotword detection. Mobvoi's speech recognition accuracy is far ahead of peers in the field of mobile search.
In some specific vertical domains, Mobvoi's semantic analysis could cover 70 verticals and comprehend difficult queries up to 11 dimensions. Unique and proprietary semantic analysis technology makes accurate mobile search possible.
Mobvoi independently develops vertical search technologies, which are based on web search architecture and supports multi-field search. Proprietary crawler system provides big data for speech, semantic analysis and proactive search. By consolidating 100+ quality data sources, Mobvoi partners with various leading mobile content providers to build the ultimate searching experiences for mobile users.
Mobvoi adopts advanced text-to-speech technologies such as parametrics speech synthesis, deep learning and unit selection to make synthesized voice accurate, smooth and natural.
Mobvoi develops a sophisticated recommendation engine to perform real-time proactive search function similar to Google Now's. Presented information such as flight, weather, train and latest news comes from a big-data based intelligent recommendation system. Personalized, proactive and practical.