Unlike last year when AlphaGo easily defeated Li Shishi, people marveled that humans were surpassed by robots. Not long ago, Ke Jie, who ranked first in the world, once again defeated by artificial intelligence (AI) robots, people have gradually gradually accepted In view of the fact that artificial intelligence surpasses humans in many fields, we will explore the opportunities for artificial intelligence-based technology to change human society from a rational perspective. According to Toutiao ’s first artificial intelligence influence report today, the Chinese national AI confidence index is as high as 83, and Chinese consumers ’attitude towards artificial intelligence has become very positive.
"In fact, artificial intelligence is gradually entering our lives, especially as a very important research on artificial intelligence speech recognition. Artificial intelligence has made very successful innovative applications worldwide, and China is also at the forefront." Xu Zhengdi, product manager of Shenzhen X-Powers Technology, a leading company in power supply and analog semiconductor technology, told the media recently. Because of the optimism about the application of voice-based artificial intelligence technology, Xinzhihui Technology has recently released the AC108, a multi-array microphone front-end ADC chip specifically for popular AI applications. "AC108, which is a very important entrance signal pick-up in intelligent voice applications, is launched. We look forward to using breakthrough SNR (signal-to-noise ratio, 108dB) performance and more optimized design to help enterprises seize market opportunities in artificial intelligence applications." Xu Zhengdi pointed out.
Figure 1: Chinese national AI confidence index reaches 83
Consumers are used to voice interaction, Smart Home is the first breakthroughThe newly released "2017 Internet Trends Report" pointed out that among smartphone users using Google Assistant (Google Voice Assistant), only 20% of mobile queries were completed by voice in May 2016, and as of May 2017 Nearly 70% of inquiries were conducted through natural / conversational language. A tremendous change has taken place in just one year, and most consumers' mobile query habits have begun to favor intelligent voice interaction.
Figure 2: Consumers are used to using voice interaction to complete queries and other operations
"Intelligent voice will first land in the vertical market. In the application scenarios where the interaction is relatively simple, the scene is relatively fixed, and the users have just needs, such as home appliance control and car electronics, there are a lot of development opportunities." Xu Zhengdi introduced in an interview. The requirements for such scenarios are relatively simple. The most needed is the recognition of a series of command words, and the technology is relatively easy to implement. At the same time, before autonomous driving matures, the driver in the car cannot release his hands, so the voice control in the car is just needed. For now, the voice control of smart homes has obviously gone a bit further. The obvious example is the popularity of Amazon Echo!
Figure 3: Voice recognition as one of the intelligent entry points of the Internet of Things
In the second half of 2016, Amazon lowered the price of the Echo Dot from $ 99 to $ 49, which means that shipments of Echo-related products will increase significantly in the next year. At the same time, the upstream supply chain revealed that Amazon has increased the number of smart speaker orders to 10 million units in 2017, three times that of 2016! In this smart speaker competition, global leaders such as Amazon, Google, and Apple are pushing more and more Internet companies to release similar smart speaker products, and many of them are Chinese companies joining the competition, such as Jingdong Intelligent's Dingdong series This has also brought huge opportunities for suppliers in various links upstream and downstream.
Signal pickup front-end processing is critical, 108dB SNR solves the biggest challenge in far-field voice designIn fact, smart voice applications at close range like Siri and Google Assistant on smartphones are relatively mature (because they are more algorithm-oriented), "and currently the most important smart voice applications usually have an interaction distance of 3 to 5 meters, or a longer distance. Far-field intelligent voice signal processing in a noisier environment is the key to Smart Home applications, which is the main target market launched by AC108. "Xu Zhengdi said," These applications require multi-array microphones and must address multi-channel voice acquisition and high signal-to-noise ratio ( SNR) and low complexity design requirements. "
The realization of intelligent voice applications requires three links: first, picking up-converting natural voice signals into digital signals; and immediately following pre-processing-performing noise reduction, echo cancellation, and dereverberation processing on the collected voice signals , To form a "clean" audio signal; finally, the local or cloud speech recognition engine recognizes and semantically analyzes the "clean" speech signal to achieve intelligent feedback. Obviously, before completing the algorithm-level speech recognition, long-distance voice pickup is required. The first thing to do is to use a microphone array and supporting speech pre-processing algorithms, such as NS, AEC, De-reverb, etc., to achieve natural language human-computer interaction.
In the entire intelligent speech system, the front-end speech collection and processing play a decisive role in the high accuracy of speech recognition. From the perspective of hardware components, the performance of the front-end voice ADC chip is the key factor! In general, one of the important parameters that determine the performance of ADC chips is the signal-to-noise ratio SNR, which characterizes the ratio of the maximum undistorted sound signal, that is, the subjectively considered useful signal strength to the noise strength, and the signal-to-noise ratio SNR The higher the noise level, the better the chip performance.
Analysis of the relevant reports of several mainstream smart speaker products on the market can find that the front-end voice ADC chip SNR is 98dB, 101dB, and the 108dB declared by AC108 is the ADC chip that achieves the highest signal-to-noise ratio among similar products in the industry. "The AC108 model name highlights its performance advantages-'Audio Codec 108dB (SNR)', 108dB is currently the highest performance for microphone array applications. At the same time, AC108 is strictly designed, produced, and tested in accordance with household appliance standards to ensure good quality requirements "" Xu Zhengdi introduced.
Figure 4: Some domestic consumer electronics products based on intelligent voice recognition technology
The whole ecological industry chain has matured, and the market is about to explodeThere is no doubt that voice intelligence applications are developing rapidly. "From the perspective of our contact with domestic and foreign companies, this market is currently facing a wave of big development opportunities. At present, major home appliance companies including Gree, Midea, Haier, and Changhong have intensively deployed intelligent voice control products, ranging from air conditioners and refrigerators. , Color TV, and even small appliances have related product layouts, some products have been launched on the market. "Xu Zhengdi pointed out," especially with the changes in cost, program maturity, and design complexity, the explosive growth of the market is just around the corner. "
According to Xu Zhengdi's analysis, in the application of large-scale household appliances such as air conditioners, the current cost of voice recognition schemes accounts for about 5% to 10% of the overall cost, and the cost pressure is still relatively large. "There is a lot of room for compression in this part of the cost. The joint ventures between Xinzhihui Technology and the domestic speech recognition industry chain are rapidly promoting low-cost, high-performance solutions." Xu Zhengdi further explained: "At present, most intelligent voice applications are SoC The standard I2S interface is generally reserved. The traditional method of supporting multiple ADC inputs is to use FPGA or MCU to implement channel conversion. AC108 considered this factor in the design, and achieved this by passing multiple voice signals through upsampling rate. Standard I2S supports multi-channel audio data transmission, which can save FPGA or MCU in the program, greatly reducing the cost and complexity of the program design. "In addition, unlike some programs that need to use multiple ADCs to support the microphone array, based on AC108 cascade two A chip (up to four cascades) can implement a multi-array microphone solution.
Figure 5: AC108 high-performance, low-cost far-field voice pickup solution
This customized optimization solution can greatly reduce the cost and accelerate the time to market of the manufacturer's products, which is especially critical for the rapidly developing consumer market. It is understood that the current core wisdom AC108 has provided potential customers with EVM Board and user-friendly supporting PC tools to facilitate engineers to quickly select ADC evaluation. "We also provide professional detailed guidance documents and reference drivers, as well as nanny-style services to help customers quickly complete design and development, and accelerate product launch." Xu Zhengdi introduced. High-performance analog signal processing is usually an engineer's nightmare. For intelligent voice recognition applications that are dominated by the consumer electronics market, such "nanny-style" technical services are very important for achieving high performance and fast time to market.
Both ADC and SoC are just one link in the speech recognition ecosystem, and the maturity of the entire ecosystem is crucial. At present, microphone array algorithms, supporting voice pre-processing algorithms, and cloud semantic platforms have matured. "A large amount of R & D work in the industry is still rapidly improving the effect of speech recognition, including multi-language support under a lot of AI training, fast and reproducible mass production multi-link cooperation, etc. The introduction of AC108 once again makes up for the high-performance ADC for microphone arrays We have already cooperated with major domestic platform solution providers and with international platforms such as Amazon Alexa. The good interaction of these industry chains is boosting the explosive growth of intelligent voice recognition. It is expected that more and more will be seen within a year. Related products are available. "Xu Zhengdi pointed out optimistically.
Disposable Vapes 3000puffs
Our products include disposable e-cigarettes, rechargeable e-cigarettes, and various of flavors of cigarette cartridges.We have our own professional team and competitive quotations for any OEM or ODM works.
We supply OEM disposable vape pen,OEM disposable electronic cigarette,ODM disposable vape pen,ODM disposable electronic cigarette,OEM/ODM vape pen e-cigarette,OEM/ODM atomizer device.
12 Flavors : Guava .Mango Ice .Passion Fruit .Mixed Fruit .Blueberry .Litchi Ice .Mango Ice .Lemon .Grapes.Watermelon .Banana Ice .Coconut
3000puffs disposable vape, Disposable E-Cigarette ,Electronic Cigarette Pod Pro,Atomizer Device Vape pod Pro Manufacturer in China
Shenzhen WeiKa Technology Co.,Ltd. , https://www.zgarpods.com