Edge AI offers an interesting future!

Artificial Intelligence or AI is becoming a common factor in almost every aspect of our lives.

In the past, putting AI to work required huge server rooms, required massive amounts of computing power, and inevitably required significant investments in energy and IT resources. Now, more tasks are being performed by devices placed at the “edge” of our physical world.

Uri Guterman, head of product and marketing at HanwhaTechwinEurope, believes edge AI will make artificial intelligence more pervasive in our world because there is no need to stream raw data back to servers for analysis. This also brings huge benefits to the video surveillance industry.

Here, Guterman explains the reasons behind this phenomenon and looks at how artificial intelligence is used today and how the technology will develop in the future.

Sustainable Development Benefits

Edge AI has several advantages over server-based AI. First, less data is transmitted back to the server, reducing bandwidth requirements and costs. The cost of ownership is reduced and there are important sustainability benefits as large server rooms no longer need to be maintained. Energy savings are also achieved on the device itself, as significantly less energy is required to perform AI tasks locally, rather than sending data back to a server.

Cost-Effectiveness

Compared with cloud-based computing models, edge AI devices typically do not require recurring subscription fees, thus avoiding the resulting price increases. Focusing on edge devices also enables end users to invest in their own infrastructure.

Greater Scalability

Cameras using edge AI can make video installations more flexible and scalable, which is particularly helpful for organizations looking to deploy projects in phases. As needs evolve, more AI cameras and devices can be added to the system without requiring end users to commit to large servers with expensive GPUs and massive amounts of bandwidth from the start.

Improving Operational Performance and Security

Since video analytics happens at the edge of the device, only metadata needs to be sent over the network, which also improves network security because no sensitive data is transmitted during transmission. The data can be intercepted by hackers. Processing occurs at the edge, so there is no need to send raw data or video streams over the network.

Because analysis is done locally on the device, edge AI eliminates the latency of communicating with the cloud or server. Faster response times mean tasks like automatically focusing on an event, granting access, or triggering an intruder alarm can happen in near real-time.

In addition, running AI on the device can improve the accuracy of triggers and reduce false positives. Through edge artificial intelligence using deep learning, people counting, occupancy measurement, queue management, etc. can all be calculated with high accuracy. This increases operator response efficiency and reduces frustration because they do not have to respond to false alarms. AI cameras can also run multiple video analytics on the same device, another efficiency gain that means operators can easily deploy AI to warn of potential emergencies or intrusions, detect security incidents or track suspects.

Video Quality Improvement

What’s more, using artificial intelligence at the edge, the quality of captured video can be improved. Noise reduction can be performed locally on the device, using artificial intelligence to specifically reduce noise around objects of interest, such as people moving within the detected area. Features like Bestshot ensure operators don't have to sift through tons of footage to find the best angle on a suspect. Instead, AI can provide the best footage immediately, helping reduce reaction times and speed up post-event investigations. It has the added benefit of saving storage space and bandwidth, since only the best photos are transferred and stored.

AI-based compression technology also applies low compression rates to objects and people detected and tracked by AI, while applying high compression rates to the remaining field of view - this minimizes network bandwidth and Data storage requirements.

Using Metadata

Edge AI cameras can provide metadata to third-party software through APIs (Application Programming Interfaces). This means that system integrators and technology partners can use it as a first means of AI classification and then perform additional processing on the classified objects with their own software – adding another layer of analysis on top of it.

Resilience

Using AI at the edge has no single point of failure. AI can continue to function even if the network or cloud service fails. Triggers can still operate locally, or be sent to another device, with records and events sent to the backend when the connection is restored.

Artificial intelligence is processed in near real-time on edge devices, rather than flowing back to servers or remote cloud services. This avoids latency analysis on potentially unstable network connections.

Benefits for Installers

For installers in particular, providing edge AI during the installation process can help them differentiate themselves in the market and provide solutions for many different use cases. An out-of-the-box solution that appeals to end users who don’t have the time or resources to manually set up video analytics.

AI cameras like the WisenetX Series and P Series work straight out of the box, eliminating the need for video analysis experts to fine-tune the analysis. The installer does not have to spend valuable time configuring complex server-side software. Of course, this also has a knock-on positive impact on training time and costs.

Looking to the future

Looking to the future, Uri Guterman said edge artificial intelligence looks very promising. More and more manufacturers are looking for ways to broaden the AI ​​camera classification and even consider AI cameras as a platform that allows system integrators and software companies to create their own AI applications that run on the cameras.

It concluded: “This is definitely an area worth exploring now for both end users and installers, as AI at the edge promises to deliver huge efficiency, accuracy and sustainability gains. ”

The above is the detailed content of Edge AI offers an interesting future!. For more information, please follow other related articles on the PHP Chinese website!