Welcome to Tencent Cloud + community, get more Tencent mass technology practice dry goods oh ~
Background: On May 23-24, Tencent’s “Cloud + Future” Summit with the theme of “Huan Qi” was held in Guangzhou. Leaders of government agencies at all levels in Guangdong Province, academic experts in the industry at home and abroad, industry giants and technology giants discussed the innovation and development of cloud computing and digital industry on the scene.
Wang Lei, director of AI application products of Tencent Cloud, made a speech entitled “Best practices of AI in traditional Industries” ** at the Cloud + Future Summit. The following content is compiled from the speech.
Wang Lei: Dear leaders, guests, partners and friends, good morning!
Just now, all the speakers shared wonderful algorithms and products. I would like to share with you how the wonderful products and algorithms are implemented. The title I share today is “AI best Practices in traditional Industries”. I’m Wang Lei from Tencent Cloud AI Application Product Center.
My share is divided into three parts. First, the superbrain, the intelligence matrix. Yesterday, Xiao Ma put forward the concept of super brain. Today, from the perspective of AI application, we look at the intelligent matrix projected by super brain, which is divided into three parts: Intelligent IssA and TI MAtrix are part super brain and part APPs, including urban super brain, including retail and medical treatment. Through such smart MAtrix, we want to provide customers with industrial super brain, provide intelligent applications and provide intelligent services together with our partners.
We provide customers with data, computing power, algorithms and scenes. In practice, we believe that scenes are very important and key. The first data comes from the scene, the second scene determines the computing power, and the third algorithm must be adapted to the scene. As we can see, our AI has been used in many scenarios, in many industries, in many scenarios for each industry, in many applications, including finance, city, transportation, commerce, health care, in fact, there are many more, education, industry, enterprise, and so on.
Let’s take a look at our practice inside the industry superbrain. Today we want to take the practice of urban super brain as an example to share with you. In the process of smart city construction, we are faced with multiple demands. The first is the demand of data, which is complicated and isolated. How can we manage and break the data island? Hardware demands, diversity, brands, manufacturers, types of diverse, how we integrate. How do we take advantage of the inconsistency between the old and the new, the software barriers, the software duplication, how do we avoid it, how do we get through it. Here we show a traditional urban monitoring system. We can see that from the lowest level of equipment to the upper monitoring center, to data extraction, to the final management system, in fact, it can be regarded as independent, which is not what we want. We upgraded to city superbrain.
From this, we can see that we have actually introduced cloud and artificial intelligence to make the whole equipment and service of the city platformized, systematized and intelligent. And we’re in the City super Brain, we emphasize ecological openness, and we welcome all partners to work with us to help our customers build their own super brains.
In the process of urban super brain construction, we will actually encounter a lot of scenes, and we would like to share with you our extraction of a scene of urban super brain. We think there are four main scenes: people, cars, situation and environment. People are very easy to understand, demography, flow statistics. In terms of cars, the previous product has been shared by students. Sentiment, public opinion, fire, danger. Environment and urban ecological environment are what we must face in smart city.
We are talking about people, cars, situations and circumstances. In the car, we put forward traffic video application. The first one is to do traffic violation detection, such as driving on the wrong side of the road, wearing a license plate, crossing the road, illegal parking, drivers’ irregular behavior, traffic congestion degree, which can do the flow optimization of the whole traffic. These are all scenarios where traffic videos can be applied.
In view of these scenes, we actually put forward the application of Youtu Skyeye Traffic. In this application, we can realize the distribution of vehicles and real-time tracking of illegal vehicles, and tracking can be carried out across the camera. That is to say, we can capture the movement track of illegal vehicles, and we have structured the pictures and videos of all vehicles, so that we can trace back afterwards. This is a very common scene, that is, car search. We can see that the system design of the tracking system actually consists of two parts. The first part is the vehicle registration database at the bottom level. On the basis of the data, we add applications, including the tracking of secret service vehicles, emergency vehicles and illegal vehicles.
Application + data finally is able to provide service, the customer need in each scenario, each platform has value, of course, here you see the main is the ability to AI, but in fact we provide service, only the AI ability is not enough, we also have a large data analysis ability, we will be more dimensional analysis was carried out on the vehicle, such as the analysis of the vehicle, The analysis of the owner, the relationship between the owner and the person. We pay attention to the vehicle’s own attributes, but also to abnormal attributes, such as behavioral attributes, such as violation information, etc. Through these big data analysis and AI technology, we finally form a good urban service application.
What I just showed you is our practice in the field of urban transportation, the service system and service application of traffic scene. In fact, in addition to traffic, we are in the public security, water, comprehensive governance, education links in each city, we and partners together, for partners to achieve intelligent application. We see so many intelligent applications, but in fact, only five basic AI technologies are used, such as vehicle detection and tracking attributes. However, these technical problems are not explained. The key technology is accuracy, which is the accuracy and weight of identification. The second is the quality of the picture, and the Angle of the shot, as you know, in practice, the Angle of the shot is very different, and the comparison of the same series.
Just see so many cities of video application, there are five basic services, our tencent AI provides 6 types of cloud, 30 children, more than 100 kinds of basic intelligence services, based on these smart services, we can build together with partner very much, very rich, very broad intelligent applications meet the needs of our customers in all walks of life.
Now, after talking about traffic intelligence applications, let’s quickly look at another example, which is our application in smart travel, smart travel we see, the traditional smart travel is the gate, and there are actually some experience problems. For example, we passengers forgot to bring the card, there is no change, the queue is very long, and very anxious, the most terrible is that there is no way to roam, the card bought in a city can only be used in this city. We want to use technology to enhance and upgrade the user experience here. We put forward the face gate, hoping to use face to change our travel experience, roaming face gate advantage is convenient, do not need to take money, we are born with the face, safe. Intelligent, truly cross – city roaming.
You can see here that we are working on the overall architecture of the requirements that support this scenario. To put it simply, it is the trinity of end, edge and cloud, responding on the end, processing on the edge and getting through on the cloud. Here we can see that in order to achieve a good travel experience, we also made a very rich system, including the gate terminal, front end, business system, as well as the identification system. In the end, what we present can not only meet the improvement of passenger experience, but also present a lot of data to operators in the way of big data. The traffic efficiency has been doubled, the rail transit standard is 15 people per minute, we can now improve many times, indicators in terms of technical indicators, we can achieve face recognition accuracy in this scene through engineering and business means, more than 3 9, the time is less than 300 milliseconds.
Just now, I would like to give you a quick introduction to our practical activities in smart city and smart travel. These practical activities are also supported by Tencent Cloud and its partners to help our customers create their own smart applications and super brains. Next, we would like to invite Mr. Zhao Dongwei, CTO of Shengzhe Technology, to share. Thank you.
Q & A How does AI empower logistics? Building a one-stop machine learning service platform, Tencent Cloud for AI to achieve more possible to build support capacity for “super brain”, Tencent Cloud focused on AI technology to land maohua: intelligent interaction, the new ecology under the help of AI
Has been authorized by the author tencent cloud + community release, the original link: https://cloud.tencent.com/developer/article/1140426?fromSource=waitui
Welcome to Tencent Cloud + community or follow the wechat public account (QcloudCommunity), the first time to get more mass technology practice dry goods oh ~