Talk to the machine and let it understand what you are talking about. Speech recognition technology turns the once-human dream into reality. Speech recognition is like a machine's auditory system, which allows a machine to transform a speech signal into a corresponding text or command by recognizing and understanding it. Since 1952, Bell Labs has developed the world's first experimental system that can recognize 10 English digital pronunciations. Today, speech recognition technology has made a breakthrough development, no longer limited to the identification of isolated words. Now, Realize continuous multi-language real-time recognition. In real life, speech recognition products have become very common, such as the Srir application software that comes with the iPhone, Google's google Now based on Android, and the very naughty Cortana speech recognition software developed by Baidu. However, what is currently popular on the market is the speech recognition realized by software. The following is an FPGA-based implementation of the language recognition board. With only one month left to fund the Indiegogo project, the fundraising of the MATRIX Voice open source voice platform has exceeded its expected target of $5000 to reach the expected 289%. MATRIX Voice is the third crowdfunding project initiated by MATRIX in Florida. In fact, the MATRIX Voice platform is a 3.14-inch ring circuit board, but this board is not only compatible with the latest language libraries but also can be used in real-time continuous languages. Identification, also recognized by a number of cloud language services, including Microsoft's recognition server, Amazon's Alexa voice service, Google's voice API and Wit's ai and Houndify. In terms of structure, MAXTRIX Voice is based on Xilinx's Spartan-6 LX4 FPGA, which can be designed as a device to plug into a low-power Raspberry Pi single-board machine or as a stand-alone board. . The two images below show the top and bottom layout views of MATRIX Voice: Figure 1 MATRIX Voice top view Figure 2 MATRIX Voice bottom view As you can see from the picture above, there are seven MEMS microphones on the back of the MATRIX Voice board, seven REG multi-color LEDs, and the Spartan-6 FPGA chip on the front panel. On the back of the AMTRIX Voice, there is a 64Mbit. SDRAM and a set of I/O interfaces compatible with Raspberry pi SBCs. Since this is the latest version of the MATRIX implementation series, there is already a more complex layered software heap support for the MATRIX Voice board, including a HAL (hardware abstraction layer) and a C++ for transparent FPGA code. The function library also has an intermediate layer between the video stream interface and the visual library (mainly the Raspberry Pi camera). In addition, it provides the MATRIX operating system and high-level API interface for the MATIX Voice top layer. At the same time, when developing based on MATRIX Voice, users can choose their favorite language according to their own habits, mainly because MATRIX Voice software library supports more languages, including mainstream C++, Python, JavaScript and Verilog and VHDL languages ​​commonly used by hardware personnel. . The following figure shows the software development level of the MATRIX Voice platform: Figure 3 Software layer view Now, the science and technology crowdfunding project has become more and more popular. The FPGA-based development function is more flexible and the frontier application is a very good idea. It not only makes the product have more free and rich interfaces, but also can be freely added. The accelerator module makes the final product not only high performance but also low cost. Xilinx has developed FPGA chips that can provide both software programming and hardware programming, so choose Xilinx to make your products more reliable and powerful. 51V Battery Pack,Portable Battery Box,Portable Battery Bank,Ac Battery Pack Zhejiang Casnovo Materials Co., Ltd. , https://www.casnovonewenergy.com
