The AI market is predicted to grow at a compound annual growth rate (CAGR) of 50% by 2025, with revenues largely focused on the use of intelligent vision technologies. As the application of deep learning in the smart home gradually deepens, including facial recognition, iris and gesture monitoring, object and free space detection will be widely used.
At present, image recognition in smart home systems is mainly face recognition, which has been widely used in several scenarios. This paper will focus on AI vision applications in home scenes, divided into three blocks: image recognition technology in home smart security, AI vision technology empowered smart home single product and development and innovation
With the development of modern science and technology and the improvement of people”s living standard, the popularity of intelligent devices has become higher and higher, and residential home intelligence will be an important development trend. In recent years, domestic and foreign research on smart home has taken shape, and the existing home monitoring system is relatively low in intelligence and cannot better adapt to the new application requirements, so it is of great importance to develop a more secure and convenient smart home system.
Artificial intelligence can act on several systems in the smart home. Among them, the visual branch of artificial intelligence – image recognition technology, involving image processing, pattern recognition, computer vision, neural network and other disciplines, is an important auxiliary tool for smart home system, which combines embedded operating system and embedded hardware platform, strengthening the combination of itself and smart home application, with new concept It has the characteristics of new concept, strong practicality and so on. At present, the image recognition in smart home system is mainly face recognition, which has been widely used in many scenarios. This paper will focus on the AI vision application in home scenario.
Image recognition technology in the home intelligent security
Image recognition is a technology that uses computers to process, analyze and understand images in order to recognize targets and pairs of different patterns. Image recognition is an important area in artificial intelligence, applied in the smart home mainly at the security level.
Smart cameras: cameras intercept images and can respond differently by recognizing the content of the image through image recognition technology. Some home cameras have machine learning technology that uses a network of neurons similar to those in the brain to adapt to new information, and can recognize if there is a dog, cat, or a delivery package at your porch. Some can home surveillance will automatically analyze recorded video, showing just a few minutes or a few screenshots to give users the information they need, rather than reviewing a full day”s worth of content. Some home monitors can recognize the difference between family members and intruders instead of triggering false alarms, and can even monitor every move of renovators during renovations, notifying users as soon as a violation occurs.
It is worth mentioning that, compared to the pixel competition of home cameras, whether they have the function of video calling, environmental awareness, object recognition, behavior recognition will gradually become the watershed of home cameras “smart” and “intellectual”, artificial intelligence technology Artificial intelligence technology can make the camera not only provide the function of shooting, discovery, identification, verification, shooting, transmission, a step in place.
Smart lock: smart lock door opening method has been following the development of biometric technology and update, press the fingerprint recognition technology is through the rapid development of fingerprint recognition module on the phone and ride on the wind. With the empowerment of AIoT, the smart lock can achieve the integration of face recognition, accurate, fast and efficient face recognition through face recognition, remote visualization and linkage defense of the smart door lock to truly achieve perception-free passage. The multifunctional alarm connected to the smart lock can be connected to the community property platform and public security system, providing users with a safe and comfortable family environment in all aspects.
However, the current face recognition smart lock faces two major challenges. One is the cost issue, face recognition module than the cost of fingerprint recognition module is too much higher, whether the product can be accepted by the public to market is still worth considering. The other is the battery life, as the face recognition industry has not been solved the problem, the smart lock is also no good solution.
Access control system: In the large system of intelligent communities, intelligent access control has become the community standard. Artificial intelligence video surveillance can achieve face recognition, vehicle analysis, video structured algorithm to extract video content, detect motion targets, classified as a variety of target information such as personnel attributes, vehicle attributes, human attributes, combined with the public security system, analysis of criminal suspect clues. At the same time artificial intelligence processing security field of massive video and monitoring will also promote the performance of artificial intelligence algorithms, and mature applications in other industries. In the smart community, including intelligent access control, vehicle gates, parking locks and other functions of the intelligent management system, can achieve the following scenarios: cell phone real name, ID card, access card binding, can be accurate personnel screening, effectively help property management.
AI identification security system has already gained application in the Pearl River Delta region of China. Various departments in Shenzhen have joined hands to launch the “snow bright” video project, that is, the installation of security monitoring and intelligent access control readers in residential buildings. At present, Shenzhen, Guangzhou, Zhongshan, Dongguan, Zhuhai, Foshan, as well as Hangzhou, Xiamen and other places are also starting the “snow bright” project.
AI vision technology widely empowered smart home single product
Today, intelligent robots in the home can identify objects and follow people through image recognition technology. With the artificial intelligence system, it can distinguish which master you are it, and can you carry out some simple interaction. For example, if it detects an elderly person in the family, it may take your blood pressure; if it is a child, it may tell you a story.
In addition, computer vision, gesture recognition and other interaction methods to become voice interaction, many manufacturers have launched with the screen speaker, and smart TV in addition to voice interaction, through computer vision analysis of video content, and content-related information for the next operation, including short video clips, while watching and buying, such as Yi equipped in the Tmall Magic Box “aim For example, Yi is equipped with the “aim” function in Tmall Magic Box. For example, in the smart refrigerator, the analysis of food in the refrigerator through computer vision and the derived functions of user health management and online shopping will be unified in the home life scene to provide a more natural interaction experience.
For women, the face recognition analysis function is more grounded. The magic mirror in the bathroom has intelligent makeup trial, facial health analysis and other functions, but also real-time records of user habits in the use of the process, to collect user preferences, combined with a large e-commerce platform, as long as consumers are satisfied with the results of the trial makeup, you can also add a key to the shopping cart to buy the desired cosmetics will soon “fly” to the consumer The desired cosmetics will soon “fly” into the hands of consumers. Intelligent magic mirror can also through the central control platform, cloud platform and smart home system, can realize the visual intercom function, and control the lights, audio, curtains, air conditioning switch. It can be said that the magic mirror is one of the smart homes with the strongest visual interaction.
Introduction to the development and innovation of AI vision technology
Currently, much research in AI focuses on the ability to analyze images from the visual system. Among them, artificial neural networks have become the most popular research area. The topology of the connections between neurons, the aggregation functions used, the threshold functions, and mathematical models such as back-propagation methods are all part of the field of AI, which we call “deep learning” and is divided into two parts – training and inference. If training is performed correctly, the neural network will provide an output response very similar to the input values of the training data set. An inference machine is a software algorithm that corresponds to the simulation of deductive reasoning. The software is usually embedded among smart home devices.
In the past few years, deep learning has been very successful in the smart home space. Image-based technologies include facial recognition, iris and gesture monitoring, object and free-space detection, and, more recently, behavior recognition. The AI market is predicted to grow at a compound annual growth rate (CAGR) of 50% by 2025, with revenues largely focused on the use of intelligent vision technologies.
The development of artificial intelligence is also dependent on the development of specialized hardware. Notably, designers and builders of vision processors, also provide software layers through embedded operating systems and/or SDKs (Software Development Kits). This makes it easy to implement software solutions and allows the hardware to maximize functionality, while also requiring platform-specific development skills that require the use of tools such as Arm”s embedOS, NVIDIA”s Jetson, Xilinx”s XSDK, and CEVA”s CDNN.
Companies developing AI for embedded systems must consider this imposed software layer when developing their solutions and designing distinctive types of hardware compatibility.
As AI continues to gain momentum, the use of AI in intelligent vision systems presents a very bright future. On the hardware side, dedicated processors have emerged; on the software side, there are increasingly powerful algorithms capable of recognizing objects, faces, and poses.
In terms of market applications of AI, the first is the smart home and smart security market; the second is mobile security systems for personal identification (unlocking, payment); and finally, biometric recognition and its application in smart buildings and smart cities. a large number of investments, acquisitions and partnerships have emerged in AI, and the market size will be considerable in the coming years, with both market and revenue Rapid growth.