This article refers to the address: http://
With the large-scale application of voice technology in navigation equipment, mobile phones, MP3/MP4, and large-scale call centers such as finance and securities, voice technology has achieved rapid development in the domestic market in recent years. Talking about this situation, Mr. Zhang Zhe, Marketing Director of Embedded Products Department of Anhui Keda Xunfei Information Technology Co., Ltd. (hereinafter referred to as Keda Xunfei) said that the real development of the domestic voice market is in the last three or four years, and at the beginning of 2000. Related products with voice technology are rarely seen in the market. One of the most important tasks of Keda Xunfei since its establishment in 1999 to 2003 is to carry out market cultivation work. Since 2004, the annual turnover has doubled. At present, it has occupied more than 60% market share of the Chinese voice technology market. Embedded voice technology products are an area of ​​great value in the future, and car navigation is an important part of its embedded voice products.The so-called voice technology is to let the intelligent machines such as computers have the technology of "can speak and listen". The two most important technologies are TTS (Text to Speech) and Speech Recognition (SR). Let the machine speak, using speech synthesis technology; let the machine understand people to speak, using speech recognition technology. Mr. Zhang Zhe said that the important value of voice technology is to improve the efficiency of human-computer interaction, making communication between people and machines as simple as communication between people. Therefore, the voice market is considered to be extremely promising. This is also an important reason for the international giants such as Google and Microsoft to invest heavily in voice-based technology and related product research. Experts also predict that in the next five years, if voice technology makes further breakthroughs, platform manufacturers, hardware manufacturers, software vendors and design companies based on this technology can form good cooperation. The market capacity of China's voice industry chain will exceed 100 billion yuan.
HKUST's voice technology and car navigation solutions
AirSound4.0 is a lightweight speech synthesis software developed by Keda Xunfei. It is small in size, low in resource occupancy and high in efficiency. It is mainly used in the speech synthesis software module in the embedded field, and is suitable for voice broadcast and application requirements in different industries.
AirSound4.0 configurable features:
Resource size configurable - minimum system size 500K
Operational efficiency is configurable - a minimum of 32 processors with a minimum memory requirement of 20MHz - Kernel requires only 32K of RAM space.
- Support multiple development platforms
-Support all Chinese character code input
- Enhanced speech synthesis
- Rich text control logo
- Powerful voice adjustment
- Support English synthesis and multilingual
- Support multiple sound effects
- a wide variety of personalized sounds
- comprehensive maintenance tools
- Support fast speaker customization service
Figure 1 AirSound basic framework
Figure 2 TTS system framework
Its embedded speech recognition product AiTalk2.0 is a high-performance embedded non-specific Chinese and English command word speech recognition engine.
The main function:
- Non-specific person identification
-Support Chinese and English recognition
- Support dynamic command addition and deletion technology features:
- Excellent platform universality
- accurate text analysis capabilities
- Fast migration capability
- Powerful domain customization
Figure 3 Identification system architecture diagram
In-vehicle navigation industry solution Keda Xunfei car navigation solution analyzes the various functions of the car navigation products and the various possible combinations of speech synthesis technology and speech recognition technology, sums up the corresponding combination of some speech functions, design principles and functions. The point chart is as follows:
Design Principles • When combined with the original car navigation function, when adding voice function, try to keep the original user interface on the car navigation product unchanged, and reduce the development workload. Speech synthesis and speech recognition are added using an additional application layer interface.
• Make minimal modifications to the hardware design and mold of the original car navigation products, and try not to increase the hardware cost.
• The content of all user voice applications can be set to let the user choose whether to turn it on or off.
HKUST Xunfei and Freescale join forces to seek a win-win situation
From the perspective of the future development of the voice market, although the prospects are extremely broad, the entire market is still in its infancy, and the manufacturers in the entire industry chain have joined hands to create a good ecological environment. This is an important factor in the development of the entire voice market and the development of the enterprise itself. key. From the perspective of voice technology and product development, more people-oriented products that can bring consumers a perfect human-computer interaction experience will be the trend of future voice technology and product design. At present, HKUST Xunfei is negotiating and cooperating with leaders in various industries to promote their development through the establishment of strategic partnerships.
In-vehicle navigation devices, which are valued by the University of Science and Technology, are growing rapidly in recent years. In 2007, shipments reached 4.1 million units, an increase of 68.2% over 2006. In the field of automotive electronics, Freescale is a global leader and its leading position is unquestionable. As the world's largest provider of automotive electronics MCUs, Freescale has the industry's most complete range of Power Architecture MCUs from 8-bit S08 to high-end 32-bit, covering all of the electronics manufacturers' needs for electronics. The introduction of Freescale's i.MX35 series of multimedia processors enables automotive OEMs to implement navigation and hands-free control of in-car radios, extending the hands-free infotainment control that was previously exclusive to luxury cars to all cars. Zhang Zhe said that they value Freescale's influence in the entire automotive electronics industry, and also value Freescale's innovative capabilities in application-based solutions.
Mr. Zhang Zhe, Marketing Director of Embedded Products at HKUST, said that Freescale's chip design is designed to meet the end user's perfect experience and to embody the features in their design specifications. For example, the Freescale i.MX35 processor allows drivers to control entertainment and navigation devices safely and easily during driving. With a simple voice command, the driver can select songs from the portable media player music set. , or get directions information anytime, anywhere. As the largest Chinese voice technology provider in China, HKUST has the leading Chinese voice core technology and Chinese voice resources.
Therefore, Mr. Zhang Zhe believes that the combination of HKUST and Freescale will produce 1+1>2 benefits. For HKUST, it will help them develop products that design more innovative applications. Freescale's resource advantages will help HKUST to better cooperate with partners and its products are more suitable for partners. Recognize, accept, and continue to maintain market leadership. For Freescale, if you can take into account the relevant Chinese speech technology elements of HKUST, you can provide differentiated products and solutions, and possibly bring more intelligent vehicles to the Chinese market. Navigation device. Of course, this will bring more user-friendly products and a more enjoyable experience to end users. Therefore, HKUST has full confidence in the prospect of cooperation with Freescale, and believes that this cooperation will not only be limited to the automotive field, but also in the multimedia and automation fields where Freescale also has advantages.
Actively developing innovative application products is the key to the next step
How to develop innovative voice technology products that are more in line with market needs and better meet the human-computer interaction experience of consumers is one of the major challenges facing the future development of HKUST. Zhang Zhe said that the University of Science and Technology has already formed two ways. It relies mainly on close communication and cooperation with partners in various industries. Internally, it has formed a mechanism in the R&D department, which is to develop the future voice technology. The direction is closely integrated with market demand. Two forward-looking speech synthesis techniques currently performed in the laboratory include emotional speech synthesis and tone conversion. The original speech synthesis products strive to achieve the naturalness and saturation of speech, while products with human emotions will be more in line with people-oriented needs. The tone conversion technology provides products with a very personal touch.
As a domestic software company, Mr. Zhang Zhe also expressed his thoughts on the future development of Chinese software companies. He believes that the core technology with independent intellectual property rights is the guarantee for the long-term development of Chinese software companies. When the international giants compete face to face, the Chinese voice core technology makes it an unbeaten position in the market. He emphasized that as a company itself, more energy should be placed on improving technological innovation. Externally, he hopes to establish a better intellectual property protection system and provide a good external environment for Chinese software companies to develop better.
About Anhui Keda Xunfei Information Technology Co., Ltd.
Anhui Keda Xunfei Information Technology Co., Ltd. was established in 1999. It is the largest Chinese speech technology provider in China. It has long been committed to the research of intelligent speech technology and has international leading in Chinese speech recognition, speech synthesis and oral evaluation. The result. At the same time, it is also one of the software companies in China that have mastered core technologies and have independent intellectual property rights. At present, it has launched a variety of voice technology products from telecommunications, finance and other industries to enterprises and home users, from PCs to mobile phones to MP3/MP4/PMP and toys to meet different application environments.
Guangzhou Yunge Tianhong Electronic Technology Co., Ltd , https://www.e-cigaretteyfactory.com