The in-vehicle voice control technology promotes the development of Internet of Vehicles

The development of human technological civilization has always been accompanied by the extension and liberation of perception and limbs. From knotting to video calls, from old horses to satellite navigation, from beacons to intelligent transportation, etc., human limbs and their perception systems are constantly expanding their scope of influence. In this process, the flow of information has also changed from simple to complex, and from one-way transmission to two-way interaction. Internet technology has enabled the networking of human thoughts, which has triggered an unprecedented information explosion in human history. As an appendage or derivative of human limbs, automobiles and their networking projects will undoubtedly bring greater changes to the expansion of human limb space. On the one hand, from the domestic hot voice navigation, people have realized that in the era of Internet of Vehicles, "moving your mouth" is much faster than "moving your hands". Following this idea of ​​navigating with your mouth, fully realizing voice control technology in the field of vehicle-mounted hosts, reforming the human-computer dialogue mode, and improving the use efficiency and user stickiness of Internet of Vehicles terminal equipment from the technical end have gradually become an important link in promoting the development of Internet of Vehicles.

Analysis of the advantages and disadvantages of voice control operation

It is difficult to say how high the contribution rate of voice control technology is to the vehicle host. In other words, if we ask which of the contact operation and non-contact operation of the vehicle host is more suitable for vehicle use, many people engaged in the vehicle host industry may suddenly wake up. At present, most vehicle hosts realize their functions through contact operations such as buttons, knobs, and touch screens. This kind of physical contact has a lot to do with the physical use limits of buttons, knobs, and touch screens, such as the damping of buttons and knobs, the life of touch screens, etc. Within the normal service life range, for a vehicle host in normal working condition, each function is accompanied by a great risk of misoperation. In terms of improving the efficiency of function control, companies led by navigation manufacturers are seeking more efficient ways. At present, a common fact in the control of vehicle hosts is that when the vehicle is in motion, if we want to change the navigation path, we need to manually touch the screen to set the destination, path, and other information. Especially when the destination is entered, the system is likely to be unable to recognize it. This situation is common in major car host functions such as radio, DVD, Bluetooth, etc. Although we have set up many shortcuts, such as navigation route memory, radio station storage, Bluetooth matching records and other personalized operating methods, in the long run, these shortcuts cannot cover everything. It is precisely because of this that there is a lack of revolutionary products in the operation of car host machines.

Due to the problem of touch sensitivity and the distance between menus, the trouble of misoperation always bothers car owners. In addition, any car host has a service life, and its buttons will become less sensitive with time and the number of applications. Once a button fails, it will take a lot of trouble to use it again.

Objectively speaking, voice control technology is a revolutionary and advanced technology for vehicle head units, but it is undeniable that voice control technology uses a non-contact operation mode after all, and its transmission mode is affected by the transmission medium. The synthesis quality during wireless transmission is not very good and needs to be further improved. The effect may be quite good in laboratory tests, but in the case of high background noise, especially when used in the car, the problem of low recognition rate arises. In addition, there are other factors that affect the performance of voice control functions.

There is no doubt that the use environment is one of the most direct reasons that affect the voice control function. Generally speaking, in a quiet environment, the voice control recognition system can show good recognition effect and has a high recognition ability, but in a noisy environment, the effect of the voice control function will continue to decline, and with the continuous increase of noise, the recognition rate will be greatly reduced, resulting in incorrect recognition. In addition, the distance of the sound source will also greatly reduce the recognition effect. The sound source here does not refer to the host, but to the human mouth. The so-called sound source distance refers to the distance between the mouth and the pickup hole of the vehicle host. Let's first learn from the voice recognition efficiency of some smart phones. Generally speaking, when people use mobile phones, they usually put their mouths to the microphone to talk. At this time, the distance between the mobile phone and the mouth is kept within 10 cm. The voice control function can play the best effect at this distance, but as the distance from the sound source continues to increase, the voice recognition rate will be significantly reduced, especially in the operating environment, which is even more unsatisfactory. It can be seen that in the car, the distance between the vehicle host and the sound source is generally about 60 cm. How to solve the energy loss of audio during transmission caused by this distance, which leads to the inability of voice control equipment to recognize, is a big problem.

The last influencing factor is attributed to the properties of the sound source. The so-called properties of the sound source should be the dialect, voice, intonation and speaking speed of the user. Different properties will have different effects on the voice control function. In terms of language, due to the obvious regional characteristics of car owners, there is usually a distinction between standard Mandarin and dialect Mandarin. The voice recognition rate of the mobile phone will be more stable in standard Mandarin, while the voice recognition rate of dialect Mandarin is relatively much worse. Of course, in terms of speaking speed and intonation, the voice control technology of different manufacturers has different effects in these aspects.

Voice control technology in vehicles

From 2010 to the end of 2011, the in-vehicle host has undergone tremendous changes in both hardware and software. The market has further unleashed its strength to move forward on the road of differentiation. From the hardware perspective, the platform of the in-vehicle host is no longer dominated by WinCE. Platforms such as Android have begun to join the in-vehicle host and continue the rapid momentum in the mobile phone field. From the software perspective, the addition of voice control technology to the in-vehicle host has become a consensus in a small range, and many manufacturers have begun mass production or pre-research.

From the current product form, the in-vehicle host that fully applies voice control technology has not yet appeared. The current in-vehicle host voice control system is only a functional selling point of the host. On the one hand, this is related to the "fence-sitting mentality" of the manufacturer. On the other hand, whether the in-vehicle host can fully realize voice control still needs to be verified by both technology and the market, and it should not be rushed. In the long run, voice control technology will be another hot spot after a round of hardware upgrades of the kernel and platform. If you look at it from this perspective, the "one-key navigation" technology that was hot in 2010 was just a "first experience" of voice control technology applied to navigation functions. Of course, the "pleasure" that this "first night" brought to the industry has not yet subsided. Unlike "one-key navigation", voice control technology embodies more of the combination of "embedded" technology and "cloud computing". Of course, this does not conflict with the services provided by "one-key navigation" in technology.

Now it seems that the biggest benefit of voice control technology in vehicles is that the car owner can control the host function without leaving his hands on the steering wheel. Perhaps, this technology can bring unexpected benefits to the high repair rate of vehicle host. Of course, the potential risks are also very high. The biggest problem is the recognition rate. How to solve the problem that the recognition rate of the vehicle host may be low under noisy driving conditions? According to the solutions that have appeared in the pre-installed and aftermarket markets, two methods may help solve this problem. One is to adopt a solution similar to the high-end version of Guangben City, which installs Bluetooth on the A-pillar on the driver's side, thus shortening the distance between the sound source and the pickup hole of the vehicle host; another solution is to take the first method a step further, and each vehicle host is directly equipped with a wireless Bluetooth, which combines the mobile phone and the vehicle host through wireless Bluetooth.

Judging from the market hot spots in 2012, voice control technology has become a new round of hot spots and it is irreversible. The market needs new technologies and new concepts to regain growth momentum. In this context, the integration and fusion trend of the vehicle host industry and related industries has also begun to emerge, such as various data exchange services based on cloud computing, application sharing on open platforms such as Android, etc. If the vehicle host was still a niche industry with a short industrial chain extension before, then with the catalysis of multiple technologies such as platforms, technologies and vehicle networking, the vehicle host is increasingly showing an open industry. Returning to the voice control technology itself, whether it can become a key link in promoting the vehicle host and vehicle networking, there are still many uncertain factors. Although many voice control related products and projects have been successful, more offline and online supporting services need to be improved.

It is precisely because of the existence of these problems of contact operation that the advantages of contactless operation are highlighted. At present, the more mature contactless operation is voice control technology. Therefore, we can also understand why the use of contactless operation will have greater advantages. The voice recognition system inside the vehicle host is a hardware device, which has stronger durability than software and peripherals (screen, buttons, etc.). Therefore, as long as you understand Chinese and pronounce it basically accurately, the voice recognition system inside the vehicle host can recognize the needs of the owner more easily. Of course, in the early stage of voice control technology being applied to vehicle hosts, there will definitely be low recognition rates, but with the emergence of competition, the corresponding voice recognition technology will definitely have a major breakthrough, thereby improving the sensitivity of voice recognition of the vehicle host.

Bottlenecks of voice control technology

The emergence of any technology, although it may have overwhelming advantages over the technology it replaces, does not mean that the technology is the best choice. This is also true for voice control technology. It has been nearly 20 years since the emergence of voice control technology and its entry into the car host industry. It cannot be said to be fast, but it is not slow at all, and at least not too late.

Based on the voice control functions and actual performance of several car head units learned from the market, the main problems are concentrated in the usage environment, and there is no problem with the technology itself.