1. Recommended Nrk2201: NRK2202-A01 module of Nine Core Electronics is an intelligent MCU module dedicated to the field of Internet of Things interaction and control. The intelligent module is based on the most advanced deep neural network (DNN-HMM) speech recognition technology at present, and realizes high recognition rate, high real-time, local and cloud integration, highly integrated speech recognition and processing functions. It can realize specific intelligent voice interaction effects such as semantic recognition.
2, CI1102 or CI1103, our single mai local speech recognition compatible module is the CI110 using neural network processor (BNPU), support speech detection, speech recognition, voice print recognition, dual Mai speech processing. Operating frequency 160MHz, support single mai far field noise reduction.
3, intelligent lighting switch manufacturers prefer nine core electronics low-cost NRK3301 offline voice chip to achieve voice control switch. NRK3301 speech recognition chip is offline speech recognition chip, using the latest neural network (TDNN) algorithm, with recognition precisionAccuracy, low error rate advantage, 5 meters far field reliable identification.
4, domestic chip leading representative Torch core technology is definitely an outstanding representative, more than 20 years of deep cultivation in the field of sound, the research of sound has an exclusive advantage, the launch of many intelligent voice platform has been adopted by many first-line brands at home and abroad, the market reputation is very good, high requirements of customers' actual combat is Torch core products over the years excellent foundation. The new intelligent voice chip platform ATS3607 (D) is recommended.
Domestic smart voice chip better smart voice chip : IFLYtek, Baidu, Alibaba, cloud sound, Jiatong Huaseng, Sibit and so on.
Local voice recognition technology Intelligent voice chip , preferred Blackwhale technology voice recognition module Intelligent voice chip ! Blackwhale based on self-developed offline speech algorithm intelligent voice chip , can output powerful voice control capabilities, including environmental noise reduction, far field voice wake up, voice recognition and other capabilities, so that fans, washing machines, range hoods and other appliances and sockets, lamps and other lightweight equipment to obtain voice control capabilities, and do not need to network.
ICRoute company's LD3320, a non-specific person's speech recognition chip, the MCU can pass the Chinese pinyin string identification words to the LD3320 by reading and writing registers, and start the speech recognition work. This individual thinks best. There are also Lingyang, Lingtong such chip manufacturers, there are also many like Chengxin Sheng Silicon JietongThe secondary development company! Not bad, too.
2, recommended nine core electronics Nrk2201: NRK2202-A01 module is dedicated to the Internet of Things interaction and control field of intelligent MCU module. It is recommended to use YQ5969, the module series can support 1-8 MI head, the voice chip can provide different solutions of single-core - 4-core 32bitARM core according to the specific needs of customers, and can support local and cloud identification of different needs. The local recognition rate within 5 meters is more than 93%, and the cloud recognition rate is 97%.
4, to see how long their voice content is, need to load how long voice, price requirements, such as ICRoute company's LD3320, non-specific person's speech recognition chip, MCU through reading and writing registers can put the Chinese Pinyin-string identification wordsSend it to LD3320 and start speech recognition. Sun microcontroller 61A can realize the voice recognition function of the chip, generally small customers can only get a specific person identification reference code.
5, recommended nine core electronics Nrk2201: NRK2202-A01 module is dedicated to the Internet of Things interaction and control field of intelligent MCU module. The intelligent module is based on the most advanced deep neural network (DNN-HMM) speech recognition technology at present, and realizes high recognition rate, high real-time, local and cloud integration, highly integrated speech recognition and processing functions. Realizable language