Xiaobu is the voice assistant of OPPO mobile phone. Xiaobu Assistant, also called Breeno, is an AI assistant built into OPPO smartphones and Internet of Things devices, including five modules: voice, advice, instruction, screen recognition and scanning. Xiaobu assistant is committed to providing multi-scene and intelligent user experience with the concept of wit, fun and warmth. Users can wake up through four entrances: voice wake-up, long press the power button for 0.5s wake-up, APP float and earphone wake-up.
Xiaobu assistant can complete simple voice communication, make phone calls, send text messages, play audio and video and other multimedia content. Function introduction: Life service. Book movie tickets, order takeout, find food, help with numbering, find recipes, send WeChat, Alipay, send red envelopes, shop, prevent and control epidemics, etc. Travel agency. Breeno Voice is the voice assistant of OPPO mobile phone. Bree has built-in functions of mainstream voice assistant software on the market.
The Development History of Voice Command
1952, Audrey, the first speech recognition system based on template matching, came out. The system can recognize isolated English numbers, and the accuracy of acquaintance recognition reaches 90%. Although the accuracy is not ideal in non-specific situations, speech recognition is recognized in public view for the first time. The National Defense Research Institute of the United States also realized the application value of speech recognition, so it increased its technical input.
197 1 year, the phonetic system for isolated words and simple sentences came out in Harpy. In the aspect of simultaneous speech synthesis technology, the design of series/parallel hybrid formant synthesizer enables the machine to synthesize more natural speech (simulated human voice).
In 1980s, the recognition of conjunctions and large vocabulary continuous speech has been greatly developed. Statistical model gradually replaces template matching method, and hidden Markov model becomes the basic model of speech recognition system. Hidden Markov model can predict on the basis of given factors by adding statistical information to signal processing technology.
At the end of 1980s, the pitch synchronous superposition time domain waveform correction (PSOLA) algorithm was proposed, which solved the problem of speech segment splicing well. The improvement of technology has triggered the initial application of voice technology, and intelligent voice has officially entered the primary stage.
The development of 200 1 deep neural network has made a qualitative breakthrough in speech technology and promoted the development of natural language processing. With the integration of semantic recognition technology, speech recognition and speech synthesis are no longer regarded as a single technology, but developed and applied as a whole speech interaction, and the concept of intelligent speech came into being.
In 2020, affected by the epidemic, the demand for intelligent voice system applications surged. Starting from the demand, intelligent voice has rapidly entered a new stage of technology integration, integrating technologies such as multi-mode recognition, big data, Internet of Things and cloud services to form a comprehensive technical solution. Multi-dimensional access to the core needs of the industry, providing more customized industry solutions has become a breakthrough for intelligent voice service providers.